Not one of them is real, and increasingly not one of them was even photographed.

Then think about the businesses in your area that people actually name when asked for a recommendation. There is a decent chance you can picture the owner's face.

That is not a coincidence. It is the entire mechanism.

In a market where any image can be generated, the one thing that cannot be faked is a specific human, appearing consistently, over time. So yes, your brand needs a face. Not a stock model, not an AI avatar, but one real person your customers see again and again until they feel they know who they are dealing with.

Why a face beats a logo now

Logos were designed for a world of scarce attention and expensive printing. They are compression, a way of packing recognition into a small mark.

They work, slowly, with enormous repetition and budget. The businesses whose logos you recognise spent decades and fortunes putting them in front of you.

A human face does the same job faster, because your brain has dedicated hardware for it. You recognise, remember and form judgements about faces in milliseconds, and you keep doing it involuntarily. A Princeton study found people form trait judgements from a face after a tenth of a second.

Which means a small business with a consistent human presence can build recognition at a speed no logo based brand can match on the same budget.

Nobody has ever felt loyal to a wordmark. People are loyal to people, and they will forgive a business a great deal if they feel they know who runs it.

None of this means you throw the logo away. The logo is the signature at the bottom of the letter. The face is the voice reading it out loud. You need both, but only one of them makes a stranger feel like they have already met you.

The AI angle, which is the urgent part

Two things happened at once.

The supply of plausible human imagery became infinite. Models, testimonials, founders, teams, all generatable in seconds. So the presence of a person in your marketing no longer proves anything by itself.

And buyers are adjusting. More of them now look for evidence that a specific, consistent, verifiable human is behind the business. Someone who appears in video with real ambient sound, in different settings, over months, saying things that connect to each other.

That continuity is the hard part to fake and the easy part to have, if you are willing to be on camera. We have written about why looking real is now the opening in a feed where everything looks generated. A consistent face is the most direct way to cash in on it.

It also does something useful in the AI answer layer. A named, consistently described human attached to a business gives models a strong entity to attach expertise to. Authorship, quotes and credentials tied to a real person are among the signals that separate a business a model will confidently name from one it will not. If you want the longer version of that argument, read how to become an entity AI trusts.

What "real" actually looks like on camera

Putting a person on screen is not enough. A glossy, over lit, heavily retouched founder portrait can read as synthetic just as fast as a generated one. The things that signal a real human are specific, and most of them cost nothing.

• Real sound. The ute door closing, the espresso machine, the wind on site. Room noise tells the viewer this happened somewhere, at some time, to someone.
• Specific details. "We rewired a 1960s weatherboard in Gosford on Tuesday" is real. "We deliver quality solutions" could be anyone, or no one.
• Imperfection you leave in. A stumble, a laugh, a dog walking through frame. Not sloppiness, just the absence of the plastic polish that now reads as fake.
• Continuity across time. The same person, with a haircut in March and a beard in June, referencing last month's job. That thread is almost impossible to fabricate convincingly for a year.
• Other real people around them. Customers, staff, suppliers. A face that exists inside a real community is more believable than one standing alone in a white void.

This is the same reason plain phone videos keep outselling polished commercials. Production value is not the enemy, but it has to serve the truth rather than cover it up.

The objections, answered honestly

"I hate being on camera." Almost everyone does, for about six weeks. Then it becomes a chore like any other. The discomfort is real and it is temporary, and it is currently one of the highest paid discomforts available to a business owner.

"What if I want to sell the business." A fair concern, and manageable. Build the personal brand as the front door and the business brand as the house. Make sure the systems, the reviews and the reputation belong to the company, and that other faces appear alongside yours over time.

"I am not the most photogenic person." Good. Photogenic reads as generated now. Ordinary reads as real, and real is the asset.

"We are a team, not a personality." Then use the team. It does not have to be the owner. It has to be someone consistent, ideally the person who actually does the work and can talk about it without a script.

How to do it without it becoming a full time job

• Pick one person and commit for twelve months. Recognition needs repetition. Rotating faces resets it every time.
• Same energy, different settings. On site, in the workshop, in the car after a job, at the desk. Variety of place, consistency of person.
• Batch the filming. One structured half day per quarter produces enough material for months. The mistake is trying to film daily and burning out in three weeks.
• Talk about the work, not about marketing. The best clips are someone explaining a problem they solved this week. That is content that only you can make and it is inexhaustible.
• Put the face everywhere it earns its keep. Ads, landing pages, email signatures, proposals, the thank you page after an enquiry. Familiarity across touchpoints is what turns recognition into trust.

The quarterly half day is where most people get stuck, so here is the running order.

1. The week before, collect questions. Pull the ten questions customers asked most in the last quarter from emails, calls and quotes. Each one is a video.
2. Write three bullet points per question. Not a script. Scripts make ordinary people sound like hostages.
3. Film the answers in one location. One camera at eye level, a clip on microphone, window light. Two or three takes each, then move on.
4. Move to a second and third setting. Film a walk through of a real job, then a few short clips in the car or on site reacting to something specific.
5. Grab stills while you are there. Twenty honest photos of the person at work will replace every stock image on your website.

A worked example

Say you run a small electrical business on the Central Coast with four staff. Your website uses stock photos of a smiling electrician who has never been to Gosford. Your ads show a lightbulb.

You pick your senior electrician, who enjoys explaining things, as the face. Every quarter you spend one morning filming twelve short answers to real questions. Why does my safety switch keep tripping. Is it worth upgrading an old switchboard. What does a rewire actually involve.

You replace every stock photo on the site with real shots of that person on real jobs. The same face goes on the quote template, the email signature and the van. Your ads switch from the lightbulb to thirty second clips of the electrician explaining one problem.

Nothing about that plan requires a production budget. It requires four mornings a year and the discipline to keep going through the quiet months at the start. The only question it leaves is whether you will still be doing it in month nine.

What to do if you genuinely cannot be the face

Some owners cannot do this. Health, privacy, the nature of the work, or a genuine and permanent hatred of cameras. That is fine, and there are two workable substitutes.

Use a consistent team member who is comfortable and who actually does the work. Customers do not need the owner. They need someone recognisable who can talk about the job with authority.

Or build recognition around a consistent visual signature instead. The same vehicle, the same uniform, the same workshop, the same hands doing the same craft, filmed the same way every time.

It works more slowly than a face. It still beats stock photography by a distance, because it is documentation of something real.

How to tell if it is working

Recognition is slow to show up in a dashboard, so track the signals that arrive first.

• Mentions in enquiries. Add "anything else we should know?" to your contact form and log every time someone references a video or names the person.
• Branded search. People searching for your business name, or the person's name plus your trade, is the cleanest sign that familiarity is building.
• Sales call length. If the first call starts with "I feel like I already know you", note it. Shorter paths to yes are the payoff.
• Watch time on the face videos versus everything else. If the human clips hold attention longer than your product graphics, you have your answer.

The compounding effect

Here is what typically happens, and the timeline is worth knowing so you do not quit in month three.

Months one to three, nothing measurable. You feel self conscious and slightly ridiculous.

Months four to six, people start mentioning your videos in enquiries. Not many. Enough to notice.

Months seven to twelve, prospects arrive already feeling like they know you. Sales conversations shorten. Price resistance softens because trust arrived before the quote did.

Year two onwards, strangers recognise you in public and referrals start coming from people you have never met, which is the strangest and best evidence that the thing is working.

None of that can be bought, generated or copied. It can only be accumulated, one uncomfortable video at a time, by a specific person who kept turning up.

That is why it is worth more every year while everything else gets cheaper to fake. If you want help planning the shoot, directing someone who hates the camera, or turning one half day into a quarter of content, that is exactly what our photo and video team does. The face has to be yours. Making it easy is our job.

Frequently asked questions

Should a small business owner be the face of their brand?

Usually, yes, if they are willing. The owner carries the most authority and the most continuity, and customers like knowing who is accountable. But it does not have to be the owner. A consistent team member who does the actual work and can talk about it naturally will build the same recognition, as long as they stay the face for at least a year.

Does using AI-generated people in ads hurt trust?

It can, because synthetic faces prove nothing. When anyone can generate a smiling customer or founder in seconds, buyers stop treating a person in an ad as evidence of anything. A real, named human who appears consistently across videos, settings and months is far harder to fake, which is exactly why it now carries more weight than a polished generated image.

How often should I post videos of myself for my business?

Consistency beats frequency. One or two short videos a week from a quarterly batch shoot is plenty for most small businesses and is sustainable for years. Filming daily usually burns people out within a month. The goal is that anyone who follows you sees the same person regularly over a long period, not a burst of content followed by silence.

What if I am not comfortable on camera?

Start with answering customer questions, because explaining your own work is far easier than performing. Use bullet points instead of a script, film several takes, and keep only the natural ones. Most people find the discomfort fades after a few sessions. If it never does, use a comfortable team member or build recognition through a consistent visual signature of your work instead.

It is not working the way it used to, and the reason has nothing to do with how well you built it.

Third party cookies, household identifiers and IP based targeting no longer deliver the scale or accuracy they once did. The signals underneath every audience you have ever built have been quietly degrading for years, and the erosion is now severe enough that precision targeting can be worse than no targeting at all.

So what replaces it? Two things. Broad audiences where your creative does the filtering, and contextual placement where you target what someone is reading or watching instead of who they are.

Which sounds like a crisis until you look at what replaced it, and realise it is the oldest idea in advertising wearing a new suit.

What actually broke

Audience targeting relied on following a person across the internet and accumulating a profile. That following happened through cookies, device identifiers and IP addresses.

Browsers restricted cookies, with Safari blocking third party cookies by default since 2020. Operating systems gave people the choice to opt out and a large share of them took it. Privacy relays hid IP addresses. Households became a mess of shared devices, work profiles and VPNs.

The result is not that targeting stopped. It is that targeting became confidently wrong. The profile still exists, it is just increasingly assembled from fragments, guesses and stale data.

Think about what that means for a carefully stacked audience. Every layer you add narrows the pool, and each layer is built on signal that may be months old or belong to someone else on the same device. You are not zooming in on your ideal customer. You are zooming in on a blurry photograph and calling it sharp.

Precision built on bad data is worse than breadth built on good creative, because precision spends your money with total confidence in the wrong direction.

Why broad plus creative now wins

Watch what the platforms did in response. They did not build better tracking. They built better prediction, and then they asked advertisers to stop narrowing the pool.

The modern model works like this. Give the system a wide audience and a diverse set of creative. The creative itself does the targeting, because who responds to which ad tells the system more about intent than any interest checkbox ever did.

A video that opens with "if your commercial fridge died on a Friday afternoon" self selects its audience in three seconds, with more accuracy than a behavioural segment built from browsing history.

This is why creative diversity became the primary lever in paid social. It is not a fashion. It is what happens when the targeting layer moves from the settings into the ad itself.

We go deeper on the case for going wide in stop building the "perfect" audience. The short version: the narrower you make the pool, the less room the system has to find the buyers your assumptions missed.

Contextual targeting, the comeback

The other half of the answer is contextual, which is targeting the content rather than the person.

Instead of chasing a specific individual across the web, you place your message against what someone is reading, watching or listening to right now. Sports coverage. Home renovation content. Business news. Parenting shows.

It sounds like a step backwards to anyone who grew up on behavioural targeting. It is not, for two reasons.

First, the content analysis available today is enormously more sophisticated than the crude category matching of a decade ago. Systems can read the actual meaning, tone and subject of a page or a video, not just its section heading.

Second, context carries intent that a profile cannot. Someone reading an article about roof leaks at 11pm is telling you something more useful than the fact that a device once visited a hardware site in March.

Context also has a quieter benefit. Your ad borrows the mood and credibility of what surrounds it. A message about business insurance sitting inside a trusted trade newsletter lands differently from the same message chasing someone around a gossip site because their cookie once visited an insurer.

What to change in your account

• Widen the targeting, deliberately. Remove the interest stacks. Keep geography, age range where it genuinely matters, and language. Let the creative do the filtering.
• Increase creative variety, not creative volume. Different hooks, formats, presenters and angles. Twelve recolours of one idea gives the system nothing to work with.
• Feed better conversion signal. Server side conversion reporting, real customer data uploaded, qualified outcomes rather than raw form fills. Prediction quality depends entirely on the quality of the examples it learns from.
• Use context where you place directly. Sponsorships, newsletters, podcasts, trade publications and connected television all let you buy the environment rather than the individual.
• Build first party data seriously. The audiences you own are the only ones not degrading, and they make everything else on this list work better.

The transition, step by step

Do not rip out every audience on a Monday morning. Move across in a way that lets you see what is working.

1. Fix the signal first. Check your pixel, set up server side conversions, and make sure the event you optimise for is a real business outcome. Our guide to event match quality covers what to check.
2. Build a creative set with real range. At least five genuinely different angles, each with its own hook, before you widen anything.
3. Run broad alongside your best existing audience. Same creative, same budget, same conversion event. Compare cost per qualified result over a few weeks, not days.
4. Shift budget toward whatever wins. In many accounts that will be broad. If it is not, look at the creative before you conclude broad does not work for you.
5. Add one contextual placement. A newsletter, podcast or publication your customers genuinely read. Give it a unique offer or link so you can track it.

A worked example: the pool builder with the perfect audience

Say you build swimming pools in Perth and spend $5,000 a month on Meta ads. Your audience is homeowners aged 35 to 55, within 30 kilometres, interested in home improvement and luxury goods, with a lookalike of past customers layered on top. It worked well a few years ago. Now costs are rising and lead quality is patchy.

You rebuild. You keep the location, drop the interests and the stacked layers, and let the creative filter. You make six ads with genuinely different angles: a family on the first hot weekend with their new pool, the real cost breakdown, the backyard that looked too small for a pool, a customer walking through the build process, the mistakes people make choosing a builder, and a myth-busting piece on running costs. Each opens with a line that only a would-be pool buyer would care about.

At the same time you set up server side conversions and upload closed jobs, so the system learns from people who actually signed rather than anyone who filled in a form. You also sponsor a local home and garden newsletter in spring. Over the following months you track cost per qualified quote, not cost per lead, and tag every ad by angle. This is a hypothetical, but it is the shape of a sensible transition. The targeting did not get cleverer. The ads and the signal did.

The part that stings

This shift takes a skill that many marketers spent a decade developing and makes it much less valuable. Audience craft. Segment building. The satisfying feeling of assembling a perfect target.

It replaces that with a harder skill. Knowing your customer well enough to write fifteen genuinely different true things to say to them, and being willing to put a real person on camera saying them.

That is uncomfortable for a lot of teams because it cannot be done in the ads manager. It requires actually talking to customers, understanding their objections, and having something to say.

And it has to keep happening. Creative that carries targeting also wears out, which is why a steady production rhythm matters so much now. We cover the cadence in your winning ad now dies in 14 days.

The measurement problem this creates

Broad targeting with creative doing the filtering produces a reporting headache worth preparing for.

You can no longer point at an audience and say that segment worked. The learning now lives in the creative, which means your reporting needs to track hooks, angles and formats rather than audience names.

Build a simple tagging habit. Every ad gets labelled with its hook type, format and angle. After a quarter you can see which arguments carry the account across every campaign, which is far more useful than knowing that a lookalike outperformed an interest stack.

That tagging discipline takes about thirty seconds per ad and it is the difference between a creative programme that compounds and one that restarts every quarter.

The upside for smaller businesses

When targeting was king, budget and data scale won. Big advertisers had more data, better lookalikes and more sophisticated segmentation.

When creative is king, the advantage moves to whoever knows their customer best and can move fastest. That is not the biggest company in your category. That is very often the smallest one, run by someone who talks to twenty customers a week.

The targeting era rewarded infrastructure. This era rewards insight and speed. If you have been losing on the first, you may be about to find you are unusually good at the second.

This is how we run paid media in our advertising work at LIQID NOISE: clean conversion signal, broad audiences, a steady stream of genuinely different creative, and reporting that tracks the ideas rather than the segments.

Go and widen your audience. Then go and make twelve genuinely different ads.

Frequently asked questions

Is interest targeting still worth using on Meta ads?

For many accounts, heavy interest stacking now narrows the pool without adding much accuracy, because the data behind those interests has degraded. Most advertisers do better keeping location, language and genuinely necessary age limits, then letting diverse creative do the filtering. Test it properly: run broad against your best interest audience with identical creative and compare cost per qualified result.

What is contextual advertising?

Contextual advertising places your ad based on the content someone is viewing, rather than a profile of who they are. An ad for roof repairs beside an article about storm damage is contextual. Modern systems can analyse the meaning and tone of a page or video, not just its category, which makes context far more precise than older keyword matching.

How does creative do the targeting in broad campaigns?

When you give an ad platform a wide audience, it learns who to show each ad to from who responds. An ad that opens with a very specific problem attracts people who have that problem, and the system then finds more people like them. That is why genuinely different hooks and angles matter more than audience settings in modern paid social.

What should I do first if my ad targeting has stopped working?

Check your conversion tracking before anything else. If the platform is receiving weak or broken signal, no targeting approach will perform. Then build a set of genuinely different creative angles, and test a broad audience against your current best audience with the same creative and budget. Judge the result on cost per qualified customer, not cost per click.

The overlap between top ranking Google pages and the sources AI assistants actually cite has fallen from around 70% to under 20%, according to Brandlight analysis cited in 5W's May 2026 research.

Read that again. Not long ago, winning Google essentially meant winning everywhere, because everything downstream drew from the same pool. Today, ranking first and being cited are two largely separate outcomes decided by two different systems.

Which means the SEO report that says you are doing well may be describing a game your customers have stopped playing.

The split is not uniform, and that matters

Not every assistant behaves the same way, and the differences tell you exactly where to spend effort.

In a June 2026 CiteLens study of 320 buyer queries, Google's AI Mode and Perplexity drew roughly 90% of their citations from Google's conventional top ten results (93% and 89%). For those two, classic SEO is still most of the battle. Rank well and you get pulled into the answer.

ChatGPT pulled only about 30% from that same pool.

That single divergence explains most of the confusion in the industry right now. A business ranking well can be highly visible in Perplexity and completely absent from ChatGPT, and both observations are correct at once.

It also explains why two people in the same office can have completely different opinions about whether "AI search" matters. The one who tests in Perplexity sees their usual Google competitors and concludes nothing much has changed. The one who tests in ChatGPT sees a different set of names entirely and concludes everything has. They are both right about the system they looked at, and both wrong about the whole picture.

Two systems, two shelves. You can be a bestseller on one and not stocked at all on the other, and the report you are reading only covers one shop.

Where the other 70% comes from

If most ChatGPT citations are not coming from the search top ten, where are they coming from?

From training data, which is a compressed impression of the whole web over years rather than a live index. From places that never ranked well but were widely discussed. Forums, community threads, industry publications, documentation, roundups, review platforms, transcripts, and long tail pages that answered something precisely.

In other words, from the part of the internet that talks about businesses rather than the part optimised to rank.

This is why a business can be invisible in search and yet turn up in an assistant's recommendation, purely because a satisfied customer wrote three paragraphs about them in a busy community thread two years ago.

Look at that list again and notice what it has in common. Almost none of it is content you publish on your own site. It is what other people publish about you, or what you publish somewhere else. A model trying to decide who to recommend is effectively asking "what does the internet say about this business when the business is not in the room?" If the honest answer is "not much," no amount of on-page polish fills that gap. That is why the reputation work below is its own strategy, not a footnote to SEO. We go deeper on one of its biggest levers in AI reads your reviews before it recommends you.

What this changes about the work

You now need two strategies running at once, and they are not the same strategy.

Strategy one, for the retrieval based systems. Conventional technical and content SEO still works and still matters. Rank for the terms tied to money. Fix your site speed, your structure and your internal linking. For Google AI Mode and Perplexity, this is roughly 90% of the job.

Strategy two, for the training and consensus based systems. This one is unfamiliar and most businesses are doing none of it.

• Be described consistently, everywhere. Same name, same services, same coverage, same specialisation across every profile, directory and mention. Contradiction is the fastest way to become the business a model is not confident enough to name.
• Get mentioned outside your own domain. Industry publications, supplier sites, association listings, roundups, podcasts, community answers. Independent corroboration is what these systems weight.
• Publish things worth quoting. Original data from your own work, specific numbers, named quotes, honest comparisons. Princeton's GEO research found that adding cited sources, statistics and quotations lifted visibility in generative engine answers materially. The practical edits are laid out in three edits that get your business quoted by ChatGPT.
• Show up where humans actually discuss your category. Not with promotional posts. With genuinely useful answers, under a real name, over time.

How to find out where you stand

You cannot fix what you have not measured, and this measurement is free.

Write down the ten questions a customer would ask before choosing a business like yours. Not keywords. Actual questions, phrased the way a person types them into a chat window.

Then ask all ten in ChatGPT, Google AI Mode, Perplexity and Gemini. Record who gets named, how they are described, and which sources are cited.

Do this once a month. It takes twenty minutes and it will tell you more about your competitive position than any rank tracker, because it is literally the answer your prospects are receiving.

Three patterns usually show up. You are absent everywhere, which is an entity and corroboration problem. You appear in the retrieval based systems only, which means your SEO is working and your reputation footprint is thin. Or you appear inconsistently, described differently in each, which means the internet holds contradictory facts about you.

Each has a different fix and you cannot tell which you have without looking.

Keep the record simple. One row per question per assistant, with columns for whether you were named, the exact words used to describe you, the competitors named, and the sources cited. After three months you will have a trend line nobody else in your market is looking at. The descriptions column is often the most revealing. If an assistant describes you with an old service, a wrong location or a specialisation you dropped years ago, that is an entity consistency problem, and our guide to becoming an entity AI trusts covers how to fix it.

The strategic read

Most businesses will keep optimising for the one system they can measure, because rank trackers produce a chart and AI visibility mostly does not yet.

That is the opening. The gap between ranking and being cited is where a smaller, sharper business can get named alongside competitors ten times its size, because reputation footprint is not proportional to marketing budget. It is proportional to how many people have said something specific about you in public.

A business with 200 detailed reviews, four industry mentions and a founder who answers questions honestly in the places customers gather will outperform a business with a bigger site and a bigger budget in exactly the systems that are growing fastest.

This work also rewards patience in a way paid media never does. An ad stops the moment the budget does. A detailed review, a genuine forum answer or a mention in an industry roundup keeps describing you for years, to people and to machines. Each one is small. Stacked over twelve months, they add up to a public record of who you are and what you are good at, and that record is exactly what the consensus based systems are reading. The businesses that start building it now will be very hard to catch later.

A worked example: two builders, two shelves

Say there are two hypothetical custom home builders on the Sunshine Coast. Builder A has invested heavily in SEO for years. Their site ranks at the top for the main local search terms. Builder B ranks on page two, but the owner answers questions in a busy regional home building forum under his real name, is listed on the local builders association site, has been featured in a regional lifestyle magazine and has a long tail of detailed reviews that describe specific projects.

Ask Perplexity or Google's AI Mode for the best custom builders in the area and Builder A is likely to appear, because those systems lean heavily on the search top ten. Ask ChatGPT and Builder B may well be named instead, because the model has absorbed years of people describing him in specific terms across sources that never ranked well.

Neither builder is doing it wrong. Each is winning on one shelf and absent from the other. The builder who wins the next few years is the one who notices and adds the missing strategy, not the one who swaps one for the other.

Do not abandon what still works

One correction worth making, because the reaction to this statistic is often too extreme.

Some businesses read that the overlap collapsed and conclude that search optimisation is finished. It is not. In the CiteLens study, two major assistants still drew roughly 90% of their citations from Google's top ten results, and ordinary Google search remains an enormous source of buying traffic.

The correct response is addition, not replacement. Keep the technical foundation, the rankings and the content that earns them. Then add the reputation footprint work that the training based systems rely on.

Businesses that abandon one for the other end up losing on both shelves at once.

Start with the twenty minute audit

Write your ten questions. Ask them in four assistants. Read what comes back about you and about the competitor who gets named most.

The difference between those two descriptions is not a mystery. It is a to do list, and it is usually shorter than you expect.

Then work it in order. Fix contradictions in how you are described first, because they undermine everything else. Earn two or three genuine mentions outside your own domain each quarter. Put numbers, named quotes and sources on your five most important pages. Keep the SEO fundamentals ticking over. Rerun the audit monthly and watch the descriptions change.

If you would rather have a team run both strategies for you, that is exactly the work our digital marketing team does: technical and content SEO for the retrieval systems, and the reputation footprint that gets you named by the rest.

Frequently asked questions

Why does my business rank on Google but not show up in ChatGPT?

Because they increasingly use different sources. In one June 2026 study, Google AI Mode and Perplexity drew roughly 90% of their citations from Google's top ten, but ChatGPT pulled only about 30% from that pool. The rest comes from training data and widely discussed sources like forums, publications, directories and reviews. Strong rankings with a thin reputation footprint outside your own site will show exactly this gap.

Is SEO still worth doing if AI answers are replacing links?

Yes. Two major assistants still draw most of their citations from conventional search results, and ordinary Google search remains a large source of buying traffic. SEO is no longer the whole job, though. Keep the technical foundation and rankings, then add reputation work such as consistent profiles, independent mentions, reviews and quotable content for the systems that rely less on search results.

How do I check if AI assistants recommend my business?

Write the ten questions a customer asks before choosing a business like yours, phrased the way people type them into a chat window. Ask all ten in ChatGPT, Google AI Mode, Perplexity and Gemini. Record whether you are named, how you are described and which sources are cited. Repeat monthly. It takes about twenty minutes and shows the answers your prospects actually receive.

Where does ChatGPT get the businesses it recommends?

Partly from search results, but mostly from elsewhere. In one June 2026 study, only about 30% of its citations came from Google's top ten. The rest reflects training data and widely discussed sources, including forums, community threads, industry publications, documentation, roundups, review platforms and transcripts. Being described consistently and specifically across those places is what makes a model confident enough to name you.

The Meta pixel audience. The Google remarketing list. The Instagram following. The TikTok audience. All of it sits on land you do not own, subject to rules you did not write and cannot appeal.

For fifteen years that was an acceptable trade, because the rent was cheap and the targeting was extraordinary.

The rent went up. The targeting got worse. Third party cookies, household identifiers and IP based targeting are far less reliable than they used to be. The signals that made rented audiences valuable have been eroding for years.

What is left is what you own, and most businesses have spent a decade not building it. The fix is simple to say: collect your own customer data with consent, keep it organised, and use it both to talk to customers directly and to feed your ad platforms better signals.

What first party data actually means

Forget the jargon. First party data is a list of people who gave you their contact details, plus what you know about them.

An email address. A phone number. What they bought, when, for how much, and what they looked at before they did.

That is it. No platform can take it away, no policy update can degrade it, and no algorithm change can put it behind a paywall.

It does three jobs, and the third is the one people miss.

It lets you reach customers directly. It lets you build far better lookalike audiences on the platforms, because you are feeding them real customer data rather than pixel guesses. And it is the raw material for feeding qualified conversions back into ad platforms, which is what stops automation optimising toward junk.

The first job is the one most people picture, and it's where email earns its reputation. We make the full case in email is the highest-ROI channel nobody respects. But treat your list as only an email list and you're using a third of the asset.

A rented audience is a tenancy. A list is an asset. One of them shows up on your balance sheet when you sell the business.

Why most lists are useless

Plenty of businesses have a list. It has 4,000 addresses, a 12% open rate, and it gets emailed when there is a sale on.

That is not an asset. That is a liability with a monthly software bill.

Three things separate a valuable list from a dead one.

How they joined. Someone who entered a competition for an iPad is worth almost nothing. Someone who downloaded your pricing guide is worth a great deal. The reason they gave you their address predicts everything that follows.

What you know beyond the address. An email with no context is a broadcast channel. An email attached to what they enquired about, what they bought and when is a targeting system.

Whether they hear from you when you want nothing. A list that only ever receives promotions trains people to ignore promotions. The businesses with lists that convert send genuinely useful things far more often than they sell.

How to build one that is worth having

• Trade something specific for the address. Not a newsletter. Nobody wants a newsletter. A pricing guide, a checklist that saves a real mistake, a comparison of the three ways to solve their problem. Specific beats generous.
• Capture at the point of highest interest. The quote page, the pricing page, the moment after a purchase. Not a footer signup nobody sees.
• Ask one qualifying question at signup. Suburb, project type, timeframe, budget range. One field. This is what turns a list into a segmented list, and segmentation is where the money is.
• Record what happens next. Enquiry, quote, sale, value. Most businesses have this in a CRM and never connect it to their marketing, which is like owning a telescope and never going outside.
• Get consent properly and honour it. Both because it is the law and because a list built on unclear consent will damage your deliverability and your reputation faster than it grows.

Notice what's missing from that list: nothing here needs a data science team. A decent form, a CRM or email platform you already pay for, and the habit of recording what happens to each lead. The hard part isn't the technology. It's deciding the list matters enough to be looked after every week.

The move that makes it pay immediately

Here is the part that turns a list from a slow burn into a performance lever this quarter.

Push your customer data back into the ad platforms. Upload your buyer list. Send qualified leads and closed revenue back as conversion events rather than raw form fills.

Two things happen. Lookalike audiences built from real buyers tend to outperform ones built from pixel behaviour. And the optimisation systems stop hunting for cheap form submissions and start hunting for people who resemble customers who actually paid you.

That single connection is the difference between automation working for you and automation working on you, and it only exists if you have first party data to send.

Here's the practical sequence:

1. Clean the list. Remove duplicates, obvious junk and anyone who has asked not to be contacted. Standardise the fields: email, phone, name, suburb.
2. Split it by value. Paying customers in one segment, high-value customers in another, open leads in a third. Each one is useful for something different.
3. Upload customer lists. Both Meta and Google let you upload customer data as audiences, matched in a privacy-protecting way. Use buyers as the seed for lookalikes, and exclude existing customers from acquisition campaigns so you stop paying to reach people who already bought.
4. Set up server-side conversion tracking. Browser-based tracking misses more events than it used to. Sending conversions from your server or CRM fills the gaps and carries better match information.
5. Send back what actually matters. When a lead becomes a qualified opportunity or a sale, report that event and its value to the platform. This is the step that teaches the algorithm what a good customer looks like.

The quality of what you send matters as much as the volume. How well the platform can match your events to real people is a number worth watching, and we explain it in the boring number quietly deciding your ad results.

A worked example: the solar installer and the form-fill trap

Say you install solar panels across regional Victoria and spend $5,000 a month on Meta ads. The campaign is optimised for leads, and it's getting plenty of them at a cost that looks great. The problem is that your sales team says most of them are tyre kickers. Renters. People outside your service area. People who wanted a free quote with no intention of buying.

The algorithm isn't broken. It's doing exactly what it was told: find the cheapest form fills. It has no idea which of those leads turned into a signed contract, because nobody told it.

Now picture the fix. You add one qualifying question to the form: do you own the home? Every lead goes into your CRM. When your team marks a lead as a booked site visit, and later as a signed job with a value, that event is sent back to Meta. You upload your last two years of actual customers and build a lookalike audience from them.

Over the following weeks, the cost per lead probably rises, because you've stopped rewarding the cheapest ones. But the leads are homeowners in your area who resemble people who've actually bought. Your sales team spends its time on real opportunities. That's the trade you want: a higher price per lead, a lower cost per customer. None of it is possible without data you own and connect.

How to tell if your data is becoming an asset

You can't manage what you don't measure, and list size on its own is the wrong scorecard. Track these instead, monthly:

• Growth from high-intent sources. How many new contacts came from pricing pages, quotes and purchases, versus giveaways and footer forms.
• Percentage of contacts with context. How many records have at least one qualifying answer and a recorded outcome.
• Engagement over time. Clicks and replies from your regular emails. A list that's warming up engages more, not less.
• Revenue traced to owned channels. Sales from email, SMS and returning customers, as a share of total revenue.
• Cost per real customer from paid ads. If connecting your data works, this should improve even if cost per lead doesn't.

The mistakes that waste a good list

• Buying or scraping contacts. They didn't ask to hear from you. They'll ignore you, complain, and damage your sender reputation and your standing with customers.
• Collecting data you never use. Twelve form fields nobody looks at annoy the customer and add privacy risk for no gain. Ask less, use all of it.
• Leaving sales data in a separate system. If your CRM and your marketing never talk, the most valuable part of your data is invisible to the tools that could use it.
• Treating consent as a checkbox. Be clear about what people are signing up for and honour unsubscribes immediately. Trust is part of the asset.
• Waiting for a perfect setup. A simple, clean list you start today beats an elaborate data platform you plan to build next year.

As targeting based on third party signals keeps getting less reliable, what you know about your own customers becomes the thing that separates your ads from everyone else's. The bigger shift behind that is covered in the targeting you trusted is quietly breaking.

The uncomfortable timeline

The businesses that will be fine in 2028 are building this in 2026. Not because a specific deadline is coming, but because lists take years to compound and there is no way to buy back the time.

A list of 300 well qualified, well segmented people who hear from you monthly and trust you is worth more than 30,000 competition entrants. It also takes a long time to build properly and cannot be shortcut.

Start now and it is a rounding error of effort each week. Start when you need it and you will be trying to build an asset in the middle of the emergency it was supposed to prevent.

The one thing to do this week

Look at your website. Find the page where people are most interested, usually pricing or services.

Put one specific, genuinely useful thing on it that requires an email address to receive. Attach one qualifying question.

Then, every month from now, email that list something worth reading that does not ask for anything.

That is the whole strategy. It is unfashionable, it is slow, and in ten years it will be the reason you still have customers when the platforms have changed shape four more times.

If you'd rather not build the plumbing yourself, that's a core part of our digital marketing work: capture points that attract the right people, CRM connections that record what happens next, and the conversion feeds that make your ad spend smarter. You bring the customers. We'll help you keep them.

Frequently asked questions

What is the difference between first party and third party data?

First party data is information your customers give you directly, such as email addresses, phone numbers, purchases and enquiries, collected with their consent. Third party data is gathered by other companies, often by tracking people across websites, and sold or rented to advertisers. First party data is more accurate, more durable and fully under your control, which is why it matters more as tracking restrictions grow.

How do I use my customer list in Meta or Google ads?

Both platforms let you upload a customer list as an audience, which they match to user accounts in a privacy-protecting way. You can use that audience as the seed for lookalike targeting, exclude existing customers from acquisition campaigns, or reach past buyers with relevant offers. Clean the list first and make sure your customers consented to marketing before uploading it.

How can a small business collect first party data?

Offer something specific and useful in exchange for an email address, such as a pricing guide, checklist or comparison, and place it on high-interest pages like pricing, services or quote forms. Ask one qualifying question at signup. Then record what happens next in a CRM: enquiry, quote, sale and value. That combination turns a basic contact list into a genuine marketing asset.

Why are my ad leads such poor quality?

Usually because the ad platform is optimising for the cheapest form submissions, since that's the only signal it receives. Fix it by sending qualified leads and closed sales back to the platform as conversion events, adding a qualifying question to your forms and building lookalikes from real buyers. The platform then learns to find people who resemble your paying customers, not just form fillers.

It can matter more to your return on ad spend than your last four creative tests combined.

It is called Event Match Quality, and if yours is below 7 you are paying a tax on every single conversion you buy. In plain terms, it scores how well Meta can connect each sale or lead you report to a real person. Better matching means the algorithm learns from accurate data, finds more buyers like your real ones, and wastes less of your budget.

Here is what it actually measures, why it got dramatically more important in the last year, and how to fix it in an afternoon.

What the score is really telling you

When someone buys from you, your website tells Meta a conversion happened. Meta then has to work out which of the billions of people it knows about that conversion belongs to.

It does that by matching the customer information you send, email, phone, name, location, browser identifiers, against its own records. Event Match Quality is a score for how confidently it can make that match.

A high score means Meta knows exactly who converted, which ad they saw, and what those people have in common. A low score means it received a conversion notification and could not confidently attach it to a human.

Now consider what that does to an optimisation system. It learns by studying the people who converted and finding more like them. Bad matching means it is learning from a blurred, partial and sometimes wrong sample.

Your creative decides who stops scrolling. Your match quality decides whether the machine ever learns why. One of those is being ignored in most accounts.

Why this got urgent

Two changes turned a technical nicety into a primary lever.

First, browser level tracking kept getting more restricted. Cookies expire faster, identifiers get stripped, and a growing share of conversions never make it back through the browser at all. What used to be automatic now has to be sent deliberately from your server.

Second, ranking systems became far more dependent on prediction. Modern delivery is essentially a forecasting engine deciding who is likely to convert before showing an ad. Forecasting engines are only as good as the training data, and your conversion signal is the training data.

So the platforms got better at using signal at exactly the moment signal got harder to collect. Accounts that solved the collection problem pulled ahead. Accounts that did not concluded the platform had stopped working.

This ties directly into how Meta now picks who sees your ads. With detailed targeting doing less of the work, the algorithm leans on your creative and your conversion data to find buyers. We cover the targeting side in why broad targeting wins and the delivery side in what Meta's Andromeda update changed. Both assume the system is being fed clean signal. Most accounts aren't.

How to find your score in five minutes

1. Open Events Manager in Meta Business Suite and select the dataset (pixel) connected to your website.
2. Go to the overview and find your key events: purchase, lead, complete registration, whichever one your campaigns optimise for.
3. Click into that event and open the Event Match Quality detail. You'll see the score and which customer parameters are being received.
4. Note which parameters are missing and what share of events include each one. This list is your to-do list.
5. Repeat for every event you optimise toward. Scores often differ a lot between events on the same site.

If you see no score at all, you're probably not sending server events yet. That's step one of the fix.

The fix, in order of impact

One. Run the Conversions API alongside the pixel. Not instead of. Both. The pixel reports from the browser, the API reports from your server, and Meta deduplicates them. Server side reporting survives everything that breaks browser tracking, which is most things.

On mainstream platforms this is a native integration or an app install rather than a development project. Check before you scope anything.

Two. Send more customer parameters. Every additional field improves matchability. Email and phone are the heavy hitters. First name, last name, city, state, postcode and country all add. Hashed, obviously, which the integration handles for you.

The single biggest wins usually come from sending information you already have. If a customer buys with an email address and your integration is only sending browser identifiers, you are withholding the one field that would have matched them.

Don't forget the identifiers that come from the ad click itself. When someone clicks your ad, Meta attaches a click identifier to the visit. If your site stores it and passes it back with the conversion, the match gets much easier. Many integrations handle this, but custom checkouts and multi-step forms often drop it along the way.

Three. Fix deduplication. If the same conversion arrives twice with different identifiers, you get inflated numbers and degraded matching. Confirm your event IDs are consistent across both channels.

Four. Clean the data at the source. Phone numbers stored in five different formats, email fields with spaces, test orders that never got excluded. Rubbish in, low score out.

What good looks like

Aim for above 7 on your primary conversion event. Above 8 is strong. Below 6, Meta is struggling to match your conversions to real people.

Check the score per event, not just overall. It is common to see a healthy score on page views and a poor one on purchases, which is precisely backwards, because the purchase event is the one that trains delivery.

And check it after every website change. Match quality quietly degrades when a form gets rebuilt, a checkout gets updated or a plugin gets replaced. Diarise a monthly look.

A worked example: the lead-gen business flying blind

Here's a hypothetical. Say you run a solar installation company in Perth, spending $6,000 a month on Meta lead campaigns. Your form sits on your website and collects name, email, phone and postcode. The pixel fires a lead event when the thank-you page loads. Nobody has ever opened Events Manager.

When you do, the lead event scores a 4. Why? The pixel only sends browser identifiers. The form collects exactly the fields Meta needs to match people, but none of them are being passed through. On top of that, a chunk of visitors use browsers that block the pixel, so some leads never get reported at all.

The fix takes an afternoon. You connect the Conversions API through your website platform's integration, pass the hashed email, phone, first name and postcode from the form with each lead event, store the click identifier from the ad, and set a shared event ID so pixel and server events deduplicate. Your test leads get excluded.

A week later, the score climbs. The campaigns haven't changed. The creative hasn't changed. But the algorithm now knows which people actually filled in the form, and it starts finding more of them. What you'd watch next: cost per lead, lead quality from your sales team, and whether your creative test results start to look more consistent. Those are the signs the machine is learning from real data.

The uncomfortable comparison

Think about where your last ten hours of marketing effort went. Probably creative. Maybe audience testing. Possibly a debate about a headline.

Now consider that an account with poor match quality is running every one of those tests on corrupted data. The winner you crowned might have won because its conversions happened to be more matchable, not because it was a better ad.

That is the real cost. Not just weaker delivery, but a testing programme producing conclusions you cannot trust, which then guide the next quarter of decisions.

Creative still matters enormously. We'd argue most ROAS problems are creative problems. But you can't diagnose a creative problem when the scoreboard itself is broken. Fix the measurement, then judge the ads.

The privacy question, answered plainly

Owners often hesitate here, and the hesitation deserves a straight answer.

Contact details such as email and phone number must be hashed before they are sent through the Conversions API. The platform matches hashes against hashes and never receives readable email addresses or phone numbers from you. Technical identifiers such as IP address and click ID are sent unhashed.

You still have obligations. Tell people in your privacy policy what you share and why, honour opt outs properly, and do not send data for people who asked you not to.

Do it properly and you get the performance without the exposure. Skipping it entirely does not make you more compliant, it just makes your advertising more expensive.

This is part of a bigger shift. Businesses that collect and respect their own customer data are in a stronger position on every platform, which is why we argue you should own your audience instead of renting it.

Do this today

Open Events Manager. Find your primary conversion event. Read the score.

If it is under 7, that number is the highest leverage thing in your marketing this month. Not the creative. Not the budget. Not the new platform someone suggested.

Fix it and the same ads, the same budget and the same offer will start performing better, because for the first time the system will actually know who is buying from you.

It is the least glamorous work in advertising and it is very close to free money. If you'd rather not wrestle with Events Manager yourself, our advertising team audits the setup, connects the Conversions API properly and keeps an eye on the score as your site changes.

Frequently asked questions

What is a good Event Match Quality score on Meta?

Aim for above 7 out of 10 on the conversion event your campaigns optimise for. Above 8 is strong. Below 6 means Meta is struggling to connect your conversions to real people, so the algorithm is learning from incomplete data. Check the score for each event individually, because a healthy page view score can hide a weak purchase or lead score.

Do I need the Conversions API if I already have the Meta pixel?

Yes, for most advertisers. The pixel reports from the browser, which is increasingly blocked or restricted. The Conversions API reports from your server, which keeps working when browser tracking fails. Running both, with matching event IDs so Meta can deduplicate them, gives the most complete and reliable picture. Many website and ecommerce platforms offer this as a built-in integration.

How do I improve Event Match Quality quickly?

Send more customer parameters with each conversion, especially hashed email and phone, which your forms or checkout usually already collect. Add name, city, postcode and country where available. Pass the ad click identifier back with conversions, fix deduplication with consistent event IDs, and clean up messy data such as inconsistent phone formats and test orders. Recheck the score about a week later.

Is sending customer data to Meta allowed under privacy law?

Contact details such as email and phone number must be hashed before they are sent through the Conversions API, so Meta receives scrambled values rather than readable emails or phone numbers. You still need to disclose this sharing in your privacy policy, honour opt-outs and avoid sending data for people who have asked you not to. If you are unsure about your specific obligations, get advice relevant to your jurisdiction.

It does not appear in your reports. It has no line in your budget. In most analytics setups it is filed under direct traffic alongside the people who typed your URL from a business card.

Which means the best performing source of visitors you have is currently invisible, unattributed, and therefore impossible to invest in.

That channel is AI assistants. People ask ChatGPT, Perplexity or Gemini who to buy from, get a recommendation, and arrive already half sold. You can measure that traffic with four layers: visible referrals, deep-page direct visits, a "how did you hear about us" field, and branded search trends.

Let us fix that, because the fix takes about an hour and it changes what you decide to spend money on for the next two years.

Why AI visitors arrive ready to buy

Think about how a normal search visit starts. Someone types a few words, gets a page of options, opens four tabs and starts comparing. You're one of the tabs. They're still shopping.

An assistant visit starts somewhere else entirely. The person has already described their problem in full sentences, asked follow-up questions, and been given a short list with reasons. By the time they click through to you, much of the comparison has already happened inside the conversation. They're not arriving to browse. They're arriving to check that what they were told is true.

That changes behaviour. Visitors who come in with a recommendation already attached tend to go deeper, read more, and act sooner, because the job of persuading them started before they landed. It's the digital version of a warm referral from a friend. And like referrals, most businesses have no idea how many they're getting.

We looked at the other side of this coin in why so many AI shoppers still bounce. The traffic is valuable, but only if your pages are built to confirm what the visitor was told.

Why AI referrals hide

Three separate mechanisms push this traffic into the wrong bucket.

Some assistants strip the referrer entirely when a user follows a link, so your analytics sees a visitor arriving from nowhere. Some route through link wrappers and redirect chains that lose the original source. And a large share of the influence never produces a click at all, because the person reads the recommendation, closes the assistant and types your business name into Google an hour later.

That last one is the biggest and the most misattributed. The assistant did the selling. Google gets the credit. Your report says organic branded search and you conclude your SEO is working.

The most valuable marketing in your business right now happens in a conversation you were not part of and cannot see. Measuring the shadow it casts is the whole job.

Layer one. Catch the visible referrals

Start with what is directly observable. In your analytics, look at referral sources and filter for the assistant hostnames.

chatgpt.com and openai.com. perplexity.ai. gemini.google.com. copilot.microsoft.com. claude.ai. Plus any category specific assistants relevant to your industry.

Whatever appears there is real, and it is the tip of the iceberg rather than the iceberg. Build it into a custom channel group so it stops being scattered across your referral report and starts being a channel you can look at on purpose.

In Google Analytics 4, the setup looks like this:

1. Open Admin, then Channel groups, and create a new group by copying the default one so nothing else breaks.
2. Add a new channel called "AI assistants." Set the condition to source matching a pattern that includes each assistant hostname above.
3. Move it above Referral in the order. Channel rules apply top to bottom, so if Referral sits first it will swallow your AI traffic.
4. Save and use the new group in your acquisition reports. It applies to historical data too, so you can see the trend immediately.
5. Add new assistants as they appear. Check your referral list monthly for hostnames you don't recognise.

Then compare that segment against everything else on three numbers. Pages per session, conversion rate, and average value. In many businesses the AI segment wins all three, sometimes by an uncomfortable margin.

Layer two. Find the hidden ones inside direct

Direct traffic is a bucket of two very different things. People who genuinely typed your homepage URL, and people whose referrer got stripped.

You can separate them with one observation. Nobody types a deep URL from memory. No human has ever manually keyed in your commercial services page followed by a hyphenated location slug.

So build a segment. Direct traffic, landing page is not the homepage. That segment is largely stripped referrals from assistants and dark social, and it is often larger than your visible AI referral segment.

Watch how it behaves. If it converts noticeably better than the site average, you have confirmed what is happening without needing perfect attribution.

Two refinements make this segment cleaner. Exclude your own team and known office IP addresses, since staff often land on deep pages directly. And exclude pages you link from email or printed QR codes, or tag those links properly, so they don't masquerade as hidden AI visits.

Layer three. Just ask

The most accurate attribution tool available to a small business costs nothing and takes ten minutes to install.

Add one optional open text field to your enquiry form. "How did you hear about us."

Not a dropdown. A dropdown gives people the options you already thought of. An open field gives you the sentence "I asked ChatGPT for someone who does this in Gosford and you came up twice."

Within a month you will have a qualitative dataset that no attribution platform can match, and you will be able to see not just that assistants are sending people, but what those people were told about you.

That second part is gold. If three prospects mention the same detail, that detail is what the machine is repeating about your business. Now you know what to reinforce and what to correct.

Ask the same question on sales calls and at checkout for phone orders. Write the answers down word for word in your CRM. Over a quarter, you'll have a record of exactly how assistants describe you, which is something most of your competitors have never seen about themselves. If you want to go further, run the prompts yourself, which we walk through in what AI is saying when customers ask about you.

Layer four. Watch branded search

Open Search Console and chart impressions and clicks for queries containing your business name, month over month.

Branded search is the exhaust of every invisible recommendation. Someone hears about you somewhere you cannot track, then searches your name. When your upstream visibility improves, this line moves before anything else does.

It is not perfect. Brand campaigns, word of mouth and offline marketing all push it too. But as a trend line, over quarters, it is the most honest signal available that your name is being said in rooms you are not in.

A worked example: the accountant who found a hidden channel

Here's a hypothetical. Say you run a small accounting practice in Parramatta. Your analytics shows most enquiries coming from organic search and direct traffic, with a sliver of referrals. You've been debating whether to cut the time you spend answering questions on forums and chasing client reviews, because it doesn't seem to produce anything.

You spend an afternoon on the four layers. The AI assistants channel group shows a small but steady stream of visits from chatgpt.com and perplexity.ai. The deep-page direct segment, visitors landing straight on your "small business tax returns Parramatta" page with no referrer, turns out to be much larger, and converts well above the site average. Within six weeks, the open text field on your form has several entries mentioning ChatGPT, and two of them say the assistant described you as "responsive and good with tradies."

Now the picture is different. The forum answers and reviews weren't doing nothing. They were feeding the recommendations that produced your best leads. Instead of cutting that work, you double down: you ask every tradie client for a detailed review and put your response time and trade experience front and centre on the landing page. That decision only happens because the channel became visible.

What to do once you can see it

Three things, in order.

• Fix the landing experience for verification. These visitors arrive to confirm a specific claim, not to browse. Lead with the fact, answer the obvious follow ups, make the next step tiny.
• Reinforce whatever the machine is saying about you. If the recommendation keeps mentioning your response time, put your response time everywhere and back it with proof.
• Invest upstream deliberately. Third party mentions, detailed reviews, roundup inclusions and extractable content on your own site. That is the machinery producing this traffic, and until now you have been running it by accident.

On that last point, two places to start: why AI reads your reviews before it recommends you, and the GEO playbook for getting cited by ChatGPT, Gemini and Perplexity.

The internal argument this wins

There is a reason worth doing this beyond curiosity.

Every business has a conversation each year where someone asks what the content, the reviews and the reputation work is actually returning. Historically the honest answer was a shrug and a story about brand.

Now you can say something better. This segment is 14% of our traffic, converts at nearly twice the site average, and has grown every month since March. That is a sentence that protects a budget.

An hour of setup, to make your best channel visible before someone decides to cut the thing that feeds it. Go and build the segment.

If you'd rather have it built for you, our digital marketing team sets up the channel groups, segments and form fields, then turns what they reveal into a plan for earning more AI recommendations.

Frequently asked questions

How do I track ChatGPT traffic in Google Analytics 4?

Create a custom channel group in GA4 Admin, add a channel for AI assistants that matches sources such as chatgpt.com, perplexity.ai, gemini.google.com, copilot.microsoft.com and claude.ai, and place it above Referral in the rule order. Then add a segment for direct traffic landing on deep pages, since much AI traffic arrives with no referrer attached and hides inside direct.

Why does AI referral traffic show up as direct traffic?

Some assistants strip the referrer when a user clicks a link, some pass visitors through redirect wrappers that lose the source, and many people never click at all. They read the recommendation, then type your name or URL later. Analytics can only record what it sees, so those visits are filed as direct or branded search instead of being credited to the assistant.

Is traffic from AI assistants worth optimising for?

For most businesses, yes. Visitors from AI assistants often arrive with a recommendation already in mind, so they tend to engage more deeply and convert more readily than general browsers. The volume may be smaller than search, but the quality is often higher. Measuring it properly is the first step, because you cannot justify investing in a channel you cannot see.

How can I find out what AI is telling people about my business?

Ask your customers. An open text "how did you hear about us" field on your enquiry form, plus the same question on sales calls, captures the exact words assistants use about you. You can also run relevant questions through ChatGPT, Perplexity and Gemini yourself. Repeated details show what the machines are saying, so you can reinforce accurate claims and correct wrong ones.

Most of it you should accept gratefully. Some of it will cost you a year of lead quality and you will not notice for six months.

The line between the two is not about how advanced the technology is. It is about one question, and once you have it, every automation decision becomes obvious.

Is this a task, or is this a judgement?

Here is the short answer. Automate the work that has a measurable right answer, like bidding, placements, resizing, first drafts, reporting and routine follow up. Keep humans in charge of the decisions that define the business, like what counts as a good customer, the offer, your brand voice, final approval and pricing. Get that split right and automation becomes an advantage instead of a slow leak.

The distinction that matters

A task has a right answer that can be measured. Which of these forty audiences responds best. What time to send. Which bid wins this auction at an acceptable cost. How to crop this image for nine placements.

Machines are extraordinary at tasks. They test faster, they never get tired, they hold no ego about which version wins, and they do not have a favourite headline they wrote on a Tuesday.

A judgement has no measurable right answer, because it depends on what you are trying to build. What kind of customer do we want. What are we refusing to do. What is a lead actually worth to us. What do we stand for. What are we willing to be disliked for.

Automation applied to a judgement does not fail loudly. It optimises confidently toward whatever proxy you happened to leave in the settings.

Automation never asks whether the target is worth hitting. It just hits it, relentlessly, for as long as you let it.

Hand over these, today

• Bidding and budget pacing. You will not beat the system at this. Stop trying.
• Placement and format adaptation. Resizing, cropping, assembling variants. Pure mechanical work.
• Audience discovery within your parameters. Broad targeting with strong creative now outperforms hand built audiences in many accounts. We make the full case in why broad targeting wins in 2026.
• First drafts of anything. Outlines, subject line options, ad variant angles, research summaries. The blank page tax is real and there is no reason to keep paying it.
• Reporting assembly. Pulling numbers into a readable summary. Nobody's judgement is expressed in a copy and paste.
• Routine follow up. Reminders, confirmations, review requests, re engagement. Consistency beats craft here.

Notice what every item on that list has in common. If the machine gets it slightly wrong, you find out quickly, the cost is small, and the fix is obvious. A badly cropped image gets fewer clicks. A weak subject line gets fewer opens. The feedback loop is short and honest. That is the signature of a task.

For the first drafts in particular, the machine is a starting point, not a finished product. There is a right way to use AI for writing without every piece sounding like it came off the same production line, which we cover in how to use AI in your marketing without sounding like a robot.

Never hand over these

What counts as a conversion. This is the most important setting in your entire marketing stack and it is usually chosen in nine seconds by whoever set the account up. If the system optimises for form submissions, it will find you an inexhaustible supply of people who submit forms. Define success as a qualified, valuable outcome and feed that back to the platform, or accept that you have outsourced the definition of a good customer to a default.

The offer. No amount of optimisation fixes a proposition nobody wants. Automation makes a weak offer fail more efficiently and at greater scale.

Brand voice and position. Generated copy converges on the average of everything ever written about your category. The average is precisely what you are trying not to sound like. Use the machine for volume and variation, keep a human on the sentences that define who you are.

Final approval on anything a customer sees. Not because the output is bad. Because the failure mode is a claim you cannot legally support, a promise you cannot deliver, or a tone that reads wrong the week something has gone wrong in your industry.

Pricing. Dynamic pricing optimises for conversion rate. Your business has to optimise for margin, positioning and the customers you want to attract. Those are different objectives.

Every item here fails the opposite way to the tasks above. When a machine gets a judgement wrong, you find out slowly, the cost compounds, and the cause is buried under months of numbers that looked fine. That delay is what makes it dangerous.

The six month failure nobody sees coming

Here is how it typically plays out, because it is rarely dramatic.

Month one, you switch on full automation. Cost per lead drops sharply. Everyone is delighted.

Month two, lead volume is up and the sales team is busy. Still delighted.

Month three, someone mentions the leads feel a bit softer. It is dismissed as a bad fortnight.

Month five, close rate has dropped hard. Revenue is flat despite double the leads. The sales team is exhausted and morale is poor.

Month six, someone finally checks and discovers the system has been optimising toward the cheapest possible form submission this whole time, and has become world class at finding people who fill in forms and never buy.

Nothing broke. It did exactly what it was told. It was told the wrong thing on day one by a dropdown nobody thought about.

The three controls that keep you safe

Feed real outcomes back. Push qualified leads and closed revenue into the platform rather than raw conversions. This one change turns automation from a risk into a genuine advantage, because now it is hunting for the thing you actually want.

The quality of that data matters as much as the choice to send it. If the platform cannot reliably match the outcomes you send back to the people who clicked, it learns from a blurry picture. We explain the number that measures this, and how to lift it, in the boring number quietly deciding your ad results.

Watch quality metrics, not volume metrics. Cost per qualified lead. Close rate. Average order value. Any of these falling while cost per lead improves is the alarm.

Keep a human on the weekly review. Thirty minutes. Read the search terms, watch the top three ads as a customer would, check where the money actually went. Automation earns you that half hour many times over, so spend it.

How to audit your automation this week

Most businesses have switched on more automation than they realise, often through defaults nobody chose. Give this an afternoon.

1. List every automated system that touches a customer. Ad campaigns, email sequences, chatbots, CRM workflows, review requests, AI writing tools. If it runs without someone pressing a button, it goes on the list.
2. Write down what each one is optimising for. Not what you hoped it would do. What the setting actually says. Cheapest lead, most clicks, most opens, fastest reply.
3. Mark each goal as a task or a judgement. "Lowest cost per click" is a task. "What counts as a lead" hides a judgement inside it.
4. Check the judgement ones against real outcomes. Pull the last three months of leads from each source and see how many became paying customers. The gaps will jump out.
5. Fix the goal before you touch anything else. Redefine the conversion, rewrite the brief, or add a human approval step. Then let the automation carry on.
6. Book the weekly review. Thirty minutes, same time every week, with a named owner. Put it in the calendar before you forget why it matters.

A worked example: the finance broker

Say you run a mortgage broking business spending $6,000 a month on lead ads, with an automated email sequence and an AI assistant drafting replies to enquiries.

The audit shows three things. The ad campaign optimises for form submissions, and the form is one tap with details pre-filled. The email sequence was written by an AI tool and never reviewed, and it promises "approval in 24 hours," which you cannot guarantee. The AI reply drafts are sent automatically after hours.

So you keep the automation and fix the judgements. The conversion becomes "a lead your team qualified after a first call," fed back to the ad platform. The form adds one qualifying question. A human rewrites the email sequence to make only claims you can stand behind. AI replies become drafts that wait for a person to approve them each morning. Bidding, placements and send times stay fully automated, because those are tasks. (The business and numbers are hypothetical, but this is the shape a typical audit takes.)

The question to ask before every handover

When you are unsure which side of the line something falls on, one question settles it.

If this runs unsupervised for six months and does exactly what I asked, will I be happy with where the business ends up.

For bidding, the answer is yes. For placement optimisation, yes. For lead volume against a poorly defined conversion event, the honest answer is absolutely not, and that tells you the setting needs fixing before the automation gets switched on rather than after.

Ask it every time. It takes four seconds and it has saved businesses entire quarters.

The right posture

The businesses that will do best here are neither the ones resisting automation nor the ones surrendering to it.

They are the ones who let the machine run the machinery, and kept a firm hand on what the business is for.

AI will not replace a good marketing operation, but it will make whatever you already have louder, good or bad. We made that argument in AI won't replace your marketing, it'll just scale what you already have, and it applies here exactly. At LIQID NOISE, our consulting work helps businesses draw this line in their own stack: what to switch on, what to fence off, and what to measure so the machine keeps hunting for the right thing.

Hand over the work. Keep the judgement. That is the whole discipline, and it is worth writing on the wall.

Frequently asked questions

What marketing tasks should a small business automate first?

Start with work that has a clear, measurable right answer and a short feedback loop. Bid management, budget pacing, resizing creative for different placements, reporting, appointment reminders and review requests are good first candidates. They save real time, and mistakes show up quickly and cheaply. Leave decisions about your offer, pricing, brand voice and what counts as a good lead with a person.

Why are my automated ad campaigns bringing in bad leads?

Usually because the campaign is optimising for the wrong conversion. If a form submission counts as success, the platform will find people who fill in forms, whether they buy or not. Redefine the conversion as a qualified lead or a sale, send that outcome back to the platform, and add a qualifying question to the form so the system learns what a good lead looks like.

Is it safe to let AI write customer emails?

It is safe for AI to draft them, but a person should approve anything a customer sees, at least until you trust the system on that specific job. The risk is not usually bad grammar. It is a claim you cannot support, a promise you cannot keep, or a tone that lands badly at the wrong moment. Use AI for speed and keep human sign-off.

How often should I review automated marketing?

Weekly, for about thirty minutes, with one named owner. Read the search terms, watch your top ads as a customer would, check where the money went and compare lead quality against lead volume. Do a deeper audit of every automated system each quarter, checking what each one is actually optimising for against the outcomes the business needs.

Search volume is falling. Buyer intent is rising.

In HubSpot's 2026 State of Marketing report, 58% of marketers said AI referral traffic carries much higher intent, with the visitors who do arrive already further along in their buying journey. Fewer visitors, warmer people.

If you judge that on a traffic chart it looks like decline. If you judge it on revenue per visitor it looks like the best year you have had.

Most businesses are looking at the traffic chart.

So is falling search traffic bad news? Not by itself. If enquiries, sales and revenue per visitor are holding or rising, you have lost the browsers and kept the buyers. The fix is to change what you measure and who you write for, not to panic about the graph.

Where the missing searches went

They did not stop existing. They moved upstream into a conversation you cannot see.

The old journey had five or six searches in it. What is this, how does it work, what does it cost, who does it near me, is this company any good, and finally the branded search that arrived at your door.

Now the first four happen inside an assistant. The buyer describes their situation in a paragraph, gets a comparison, asks two follow ups and forms a view. Then they search, once, for the business they have already half chosen.

So you lost four searches and kept the one that was worth having.

Google is part of this too. When the results page itself answers the question, the click never happens. We cover how that plays out in Google just ate your traffic. Either way, the research is still happening. It is just happening somewhere your analytics cannot see.

You did not lose traffic. You lost the browsing. What is left is the buying, and it costs the same to serve while being worth several times more.

Why this breaks your reporting

Every standard marketing report is built on volume. Sessions, impressions, clicks, followers, reach. All of it assumes that more attention arriving at the top produces more sales at the bottom, at a stable rate.

That assumption just broke. The rate changed. Dramatically.

Which produces a specific and painful failure. Traffic halves, enquiries hold steady, revenue goes up, and the marketing budget gets cut anyway because the graph everyone looks at in the meeting is pointing down.

I have watched businesses cancel the channel that was quietly closing their best work because the dashboard made it look sick.

Volume numbers were always a proxy. They were a decent one while the conversion rate stayed stable. Now that it does not, a volume report is closer to a vanity metric than a business metric, the trap we pull apart in vanity metrics are quietly bankrupting your marketing.

The numbers to run instead

Four metrics, all available to you today, no new tools required.

• Revenue per session. The single most clarifying number in the new environment. If sessions fell 40% and revenue per session doubled, you are winning and the traffic chart is lying.
• Enquiry to sale rate. When buyers arrive further along, this rises. A jump here is the fingerprint of pre educated demand, and it is the strongest evidence that your invisible upstream presence is working.
• Branded search volume. When an assistant recommends you, this is where it lands. Track it monthly in Search Console. It is the closest thing to a scoreboard for AI visibility that exists right now.
• Sales cycle length. Better informed buyers decide faster. If your average time from enquiry to close is shortening, the upstream work is doing its job even though no report shows it.

How to set these up in an afternoon

1. Revenue per session: export monthly sessions from your analytics and monthly revenue from your accounts. Divide one by the other in a spreadsheet. If you sell offline, use enquiries per session as a stand-in.
2. Enquiry to sale rate: count enquiries and closed sales per month from your CRM or inbox. Even a rough count works.
3. Branded search: in Search Console, filter queries containing your business name and record total impressions and clicks each month.
4. Sales cycle: for each closed sale, note the date of first contact and the date of purchase. Average the gap monthly.
5. Add one question: ask every new enquiry how they found you, with "an AI assistant like ChatGPT" as a named option.

That last one matters more each month, because assistant referrals often show up in analytics as direct or branded traffic. The only reliable way to catch them is to ask. There is more on tracking that traffic in traffic from AI converts better and almost nobody is tracking it.

What to change about the work itself

Stop producing top of funnel explainers. The "what is" and "how does it work" content that used to bring the volume is now answered before anyone arrives. Writing more of it is producing supply for a demand that moved.

Write for the person who already understands. Comparisons. Selection criteria. What goes wrong. Real pricing. When you are the wrong choice. These pages survive because they require judgement rather than a definition, and they meet the buyer where they actually are now.

Make your sales process match the buyer. This one gets missed constantly. If prospects arrive having done four rounds of research, a discovery call that starts with "so tell me a bit about your business" wastes the advantage. Skip ahead. Ask what they have already ruled out and why.

Raise your standard for what counts as a lead. With fewer, better prospects, the cost of chasing an unqualified one goes up in relative terms. Tighten the form. Ask a qualifying question. Fewer, better is now the entire model.

A worked example: the accounting firm with the falling graph

Say you run a small accounting practice in Canberra. Your monthly report shows website sessions sliding steadily over the past year. Your blog posts explaining "what is a BAS" and "how does GST work" used to pull in most of your traffic, and they have dropped hardest. Your business partner wants to cut the SEO budget.

Before anyone decides, you run the four numbers. Imagine enquiries have stayed roughly flat, the share of enquiries that become clients has gone up, more new clients mention they "asked ChatGPT for a local accountant", and branded searches for the firm's name have crept up. Revenue per session has climbed because the visitors who remain are the ones ready to engage an accountant.

The decision flips. Instead of cutting, you stop writing beginner explainers and rewrite your three key service pages for business owners comparing firms: who you are best for, what a typical engagement includes, your fee ranges and what changes them, and when a different kind of firm would suit them better. You change the first call to start with "what have you already looked at?" This is a hypothetical, but it shows how the same data can justify opposite decisions depending on which line you look at.

The strategic read

There is a version of this shift that should genuinely cheer up any business with real expertise.

The old game rewarded whoever could produce the most content and buy the most clicks. Volume beat quality with depressing reliability, because volume caught people early and nurtured them into buyers.

The new game removes the early stage from your hands entirely. What is left is the part where a human decides who to trust with something that matters. That decision is made on evidence, specificity, reputation and clarity, which is exactly the ground a good small business can win on.

The content mill cannot compete for a buyer who has already read forty summaries and now wants to know who has actually done this before.

The pricing conversation nobody expects

One more consequence worth planning for, because it catches businesses off guard.

Buyers who arrive after four rounds of AI research often arrive knowing the price range in your category. They have been told what things typically cost, sometimes accurately and sometimes not.

That kills the old approach of holding price back until you have built value in a meeting. The value has to be built before the conversation, in the material a machine can read.

Publish your pricing structure, or at least your ranges and what drives them. It filters out the wrong prospects, it gives assistants accurate information to repeat, and it means the people who do reach you have already accepted the number.

If you are nervous about publishing prices, start with what makes a job cost more or less. That alone answers most of the questions a buyer brings, and it positions you as the honest option in a category where competitors hide behind "contact us for a quote".

What to do this month

Pull twelve months of data. Chart sessions and revenue on the same page. Then chart revenue per session on its own.

If that third line is going up while the first is going down, print it and take it to whoever is worried about the traffic chart.

Then take the three pages that generate your actual enquiries and rewrite them for someone who already knows what they want and is deciding between you and two others.

That is where every remaining visitor now is.

If you want help rebuilding the reporting and the pages, that is exactly what our digital marketing and AI search team does at LIQID NOISE. We set up the numbers that show what is really happening, then write for the buyer who has already done the research.

Frequently asked questions

Why is my website traffic dropping but my sales are not?

Many early-stage questions are now answered by AI assistants and search result summaries before anyone clicks. The people who still visit tend to be further along the buying journey, so fewer visits can produce the same or more sales. Check revenue per session and your enquiry to sale rate to see whether the traffic you kept is simply worth more.

What should I measure instead of website traffic?

Track revenue per session, enquiry to sale rate, branded search volume and sales cycle length. Together they show whether visitors are arriving better informed and more ready to buy. Add a "how did you find us?" question to every enquiry, including AI assistants as an option, because those referrals often hide inside direct or branded traffic in analytics.

Should I still write blog posts if search volume is falling?

Yes, but change what you write. Basic "what is" explainers are increasingly answered before anyone reaches your site. Focus on content for buyers who already understand the topic: comparisons, selection criteria, real pricing, common failures and honest guidance on when you are not the right fit. These pages need judgement, which keeps them valuable to readers and to AI systems.

Should small businesses publish their prices online?

In most cases, publishing price ranges and what drives them helps. Buyers often arrive already knowing typical costs from AI research, so hidden pricing feels evasive. Published ranges filter out poor-fit prospects, give AI assistants accurate figures to repeat, and mean the people who contact you have already accepted roughly what you charge.

Marketers themselves estimate they waste around a quarter of their budgets. In a Rakuten Marketing survey of 1,000 marketers, the average estimate was 26%, lost to ineffective channels and strategies. That is the people running the campaigns saying it about their own work. The real figure in an unaudited account can easily be worse.

On a two thousand dollar a month budget, a quarter is six thousand dollars a year, gone, for nothing. On ten thousand a month it is thirty thousand.

So where does it actually go? Mostly into seven leaks: branded clicks you would have got for free, broken tracking, campaigns bidding against each other, irrelevant search terms, landing pages that lose the click, tired creative and ads shown to people who already bought. Every one of them can be found and fixed without a bigger budget.

Here is the audit. It takes about two hours and it usually pays for the year.

One. Branded search you already owned

Start here because it is the biggest and the least discussed.

Open your search campaigns and split performance into branded and non branded. Branded means anyone searching your business name or a close variant.

Those conversions look magnificent. Low cost, high rate, excellent return. They are also, mostly, people who were coming to you anyway and would have clicked the organic result sitting directly underneath.

Automated campaigns love this traffic because it makes them look good. Left alone, they will drift more and more budget toward it. That drift is getting harder to see as platforms push advertisers into broader automated campaign types, which is exactly why we flagged it in Google moving campaigns to AI Max.

You may still choose to buy some branded traffic, for competitor defence or for control of the message. Make it a decision rather than a default, and never let branded results inflate the numbers you use to judge everything else.

How to do it in practice:

1. Build a list of your brand terms, including common misspellings and your founder's name if people search it.
2. Add them as negatives in your non branded campaigns so the two never mix.
3. Run branded search in its own small campaign with its own budget cap.
4. Report the two separately, every time, so a strong brand week never hides a weak prospecting week.

Two. Tracking that was broken before you noticed

A lot of "our ads stopped working" conversations end at a broken pixel.

Check three things. Are conversions being counted more than once, for example a form submission that also fires on the thank you page. Are you counting non events like phone number clicks or PDF downloads as if they were sales. And is your conversions API running alongside your pixel with an event match quality above 7.

Bad signal does not just mislead you. It actively misdirects the algorithm, which spends real money finding more of the wrong thing.

Every automated ad system is a mirror. Feed it a distorted picture of success and it will go out and buy you more distortion, faster and cheaper every month.

The quickest test is to compare what the ad platform says against what actually happened. Pull last month's conversions from the ad account, then count the real enquiries in your inbox or the real orders in your store. If the platform reports far more than reality, something is double counting. If it reports far fewer, something is not firing. Either way, you have been optimising on fiction. If match quality is the weak spot, the boring number quietly deciding your ad results walks through fixing it.

Three. Audience and campaign overlap

Older accounts accumulate campaigns like a shed accumulates paint tins. Six ad sets, four of them targeting broadly overlapping people, all bidding in the same auction.

You are, in a real sense, competing against yourself and paying the increment.

Consolidate. Modern platforms want fewer, better fed campaigns. Most accounts should be running the majority of spend through a small number of consolidated campaigns rather than a museum of legacy structures each holding a fraction of the data.

The symptom to look for is lots of ad sets each spending a small amount, none of them getting enough conversions to learn properly. Pick your strongest one or two, move the budget into them, and pause the rest. It feels like losing control. In practice it gives the algorithm enough data to do its job.

Four. The search terms nobody has read

Open the search terms report. Sort by spend. Read the top fifty.

In almost every account there will be terms that make you wince. People looking for jobs. People looking for DIY instructions. People looking for a competitor's warranty department. People looking for a service you stopped offering in 2022.

Broad matching and automated targeting pull all of it in. Some of it is genuinely valuable and you would never have thought to target it. The rest is money set on fire in small denominations.

Build the negative list. Then diarise this for every fortnight, because it refills.

Start the list with the usual suspects: jobs, careers, salary, free, DIY, how to, course, training, and the names of competitors you do not want to bid on. Then add anything specific to your industry that showed up in the report. And when you find a great term you had never thought of, add it as a keyword in its own right so you can bid on it deliberately.

Five. Traffic sent to a page that leaks

This is the waste that does not appear in the ad account at all, which is why it survives so long.

You can have perfect targeting, excellent creative and a sensible bid, then send the click to a homepage with nine navigation links, a slideshow and a form below the fold that asks for eleven fields.

Check three numbers per landing page. How long the page takes to become usable on a phone on a normal connection. What percentage of visitors start the form. What percentage finish it.

A form that loses 60% of the people who start it is not a form problem you can outspend.

The fixes are rarely glamorous. Send each campaign to a page that matches the ad's promise, word for word where you can. Strip out the navigation. Put the form or the call button where a thumb can reach it without scrolling. Cut every field you do not need to make first contact. Put proof, reviews and guarantees next to the button. If you want the full blueprint, see the anatomy of a landing page that prints money.

Six. Creative that finished working weeks ago

Winning ads can tire within a few weeks now. If your best performing ad launched in June and it is now September, you are not running a winner. You are paying a premium to show a tired ad to people who have already decided.

Check the launch date on every ad currently taking real budget. Anything older than a month with rising frequency and falling click through is quietly taxing you.

The answer is not to panic and rebuild everything every fortnight. It is to keep a pipeline. Have new hooks and new angles in production before the current winners fade, so a replacement is ready the week performance dips. We covered the mechanics in your winning ad now dies in 14 days.

Seven. Retargeting people who already bought

Small, embarrassing, and present in most accounts. Exclude converters from prospecting and from most retargeting. Otherwise you are paying to advertise a product to the people who already own it, which is both wasteful and mildly insulting to a new customer.

The fix is to upload or sync your customer list and build an audience of recent purchasers, then exclude it wherever you are trying to win new business. Keep a separate, deliberate campaign for existing customers if you have something genuinely new to offer them, like a refill, an upgrade or a complementary product. That is marketing. Showing them the ad for the thing they bought last week is not.

The two hour version

If you only have one sitting, do it in this order.

• Split branded from non branded search and judge them separately.
• Verify conversion tracking, remove duplicates and non events.
• Read the top fifty search terms by spend and build negatives.
• Check launch dates on every ad taking meaningful budget.
• Load your main landing page on a phone and try to submit the form yourself.
• Confirm converters are excluded from prospecting.

Plenty of businesses find a real share of their spend sitting in those six checks. Not through incompetence. Through accumulation, which is what happens to every account nobody audits.

Write down what you find as a dollar figure, not a vague note. "Branded search took a third of the budget" is easy to ignore. "We spent this much last quarter on people who were already typing our name" gets acted on.

A worked example: the physio clinic audit

Say you run a physio clinic in Hobart spending $3,000 a month across Google and Meta, and your ad agency reports a healthy return every month.

You sit down for two hours. The branded split shows a big share of your Google conversions come from people searching your clinic's name. The search terms report is full of people looking for physio jobs and free exercise sheets. Your Meta pixel is firing a lead on both the form submit and the thank you page, so every enquiry is counted twice. Your best ad has been running since autumn. And your booking link sends people to a homepage with a slideshow and a long form.

None of that required new strategy to find. Fix the tracking, cap branded spend, add negatives, refresh the creative and point the ads at a proper booking page, and you have the same $3,000 doing noticeably more work. The exact figure will depend on your account, which is the point of auditing it rather than guessing.

Then do the harder thing

Once the waste is out, resist the urge to bank it. Move it into the two things that compound rather than evaporate. More genuinely distinct creative, and a better offer.

Recovered waste spent on the same weak proposition just buys you more of the same result, slightly cheaper. Recovered waste spent on something worth advertising changes the shape of the year.

And make the audit a habit, not an event. Search terms every fortnight, creative dates monthly, tracking and structure every quarter. Waste comes back the moment you stop looking. If you would rather have someone else do the looking, our advertising team runs this exact audit on Google and Meta accounts, then rebuilds the parts that leak. Either way, open the account this week. The money is already going somewhere.

Frequently asked questions

How much ad spend is typically wasted?

In a Rakuten Marketing survey of 1,000 marketers, respondents estimated they waste an average of 26% of their budgets, and accounts that have never been audited can be worse. The waste rarely comes from one big mistake. It builds up from branded clicks you would have got anyway, broken tracking, irrelevant search terms, overlapping campaigns, weak landing pages, tired creative and ads shown to existing customers.

Should I stop bidding on my own brand name?

Not necessarily. Bidding on your brand can protect you when competitors bid on your name and lets you control the message. The problem is letting branded results mix with everything else, which makes your campaigns look far better than they are. Run branded search separately with its own budget cap and judge it on its own numbers.

How often should I audit my Google and Meta ad accounts?

Review search terms every fortnight, since irrelevant queries keep appearing. Check creative performance and launch dates monthly, because ads fatigue quickly. Do a full audit of tracking, campaign structure, exclusions and landing pages at least once a quarter, and any time results suddenly change. Regular small checks stop waste building up again.

How do I know if my conversion tracking is broken?

Compare the conversions your ad platform reports with what really happened, such as enquiries in your inbox or orders in your store, for the same period. Big gaps in either direction point to a problem. Also check whether one action fires multiple conversion events, and whether small actions like button clicks are being counted as if they were sales.

ACP. UCP. AP2. Visa Trusted Agent. Mastercard Agent Pay.

You do not need to love them. You do need to know what they are, because a merchant that supports only one of these protocols can be invisible to agents built on the others.

This is the plumbing layer of agentic commerce, and it is one of those rare moments where a boring technical decision has an obvious commercial consequence.

Put simply, these protocols are the rules that let an AI agent prove who it is, read your product data and pay you. If agents can't do those three things on your site, they buy somewhere else.

The problem all five are solving

An AI agent trying to buy something on your behalf hits three walls.

It cannot prove it is a legitimate agent acting for a real person rather than a bot scraping your inventory. It cannot reliably read your product data, because every site structures it differently. And it cannot pay, because payment systems were built on the assumption that a human is present with a card.

Each protocol is an attempt to solve one or more of those. Identity, data and payment.

It's worth sitting with how deep that goes. Almost every part of a normal online store assumes a person is on the other end. Product pages are designed for eyes. Checkout flows are designed for fingers. Fraud systems are designed to flag anything that doesn't behave like a human. An agent trips over all of it, which is why a new layer of rules is being built underneath.

The last twenty years of ecommerce assumed a human with a browser. The next ten assume a machine with a mandate. Almost none of the infrastructure transfers.

The five, in plain English

ACP, the Agentic Commerce Protocol. The most visible one, driven by OpenAI's shopping surfaces. Its current shape is discovery inside the assistant and checkout in the merchant's own environment. The agent finds the product, presents it, then hands the buyer to you to complete. Note that the earlier instant checkout model was retired in March 2026 in favour of this. Discovery happens in the assistant, the transaction happens on your site.

UCP, the Universal Commerce Protocol. A broader attempt at standardising how product, availability and pricing data are exposed to any agent, rather than to one platform's shopping surface.

AP2, the Agent Payments Protocol. Focused on the payment problem specifically. How does a merchant know an agent has genuine authority to spend a human's money, and who carries the liability when it goes wrong.

Visa Trusted Agent and Mastercard Agent Pay. The card networks answering the same question from their side. Both establish a way to verify that an agent is authorised, so a transaction initiated by software is not automatically treated as fraud.

The overlap is messy and the standards are still moving. That is normal for this stage and it is not a reason to wait.

If you want the simplest mental model, map each one to the three walls. ACP and UCP are mostly about data and discovery: can the agent find and understand what you sell. AP2 and the card network programs are mostly about identity and payment: can the agent prove it's allowed to buy, and can you trust the transaction. A business that is readable, reachable and payable is covered on all three.

Where the traffic is actually going

The share shift is worth understanding, because it tells you the game is not yet decided.

Similarweb data reported by Digiday showed ChatGPT made up less than 3% of Amazon's referral traffic in August 2025, down nearly 18% on the month, after Amazon blocked many AI crawlers. The same month ChatGPT drove about one in five of Walmart's referral clicks, more than 20% of Etsy's, nearly 15% of Target's and 10% of eBay's. Shopify says millions of its merchants can now sell to ChatGPT users, and both Walmart and Target have announced partnerships with OpenAI.

Read that carefully. The dominant player in ecommerce search is not the dominant player in agentic commerce. When a discovery layer changes, the incumbents do not automatically carry their position across.

That is the whole opportunity for a smaller business. The shelf is being rebuilt and nobody has been given their old position back. The bigger picture of how buyers are handing their research to machines is laid out in AI agents are shopping for your customers.

What to do, in the order that matters

Most businesses do not need to implement a protocol themselves. They need to be readable, then supported, then integrated. In that order.

• Fix your structured product data first. Clean schema markup on every product and service. Price, availability, specifications, shipping, returns, service area. This benefits every protocol and works today regardless of which standards win.
• Check what your platform already supports. If you are on Shopify or a major platform, a lot of this is a settings toggle and a feed, not a development project. Find out before you scope anything.
• Make sure agents can reach you. Aggressive bot blocking, interstitials, and login walls will exclude you from agentic traffic entirely. Review what your firewall and CDN are actually blocking. Plenty of businesses have quietly banned the exact traffic they are trying to attract.
• Write returns, warranty and shipping terms as plain readable text. Agents compare on these constantly and cannot read them out of a PDF or an image.
• Then support more than one protocol. Every protocol you support makes you reachable by more agents. Do not bet on a single standard this early.

On the reachability point, here's a practical way to check it. Ask whoever manages your site to pull up the bot protection settings in your firewall or CDN and list which categories of automated traffic are blocked or challenged. Then check your robots file for rules that shut out AI crawlers. Many security tools ship with broad bot blocking switched on by default. That was sensible when all bots were scrapers. It isn't any more, and the decision about which agents to allow should be a deliberate business call, not a default nobody reviewed.

A worked example: the online store that was invisible to agents

Say you run an online store in Perth selling outdoor furniture. Good products, decent Google rankings, a loyal customer base. You've heard about AI shopping and assume you're covered because you're on a major ecommerce platform.

Then you actually check. Your product pages have prices, but dimensions and materials sit inside an image of a spec sheet. Delivery costs only appear at checkout. Your returns policy is a downloadable PDF. Stock levels aren't marked up, so an agent can't tell if the table is available. And your CDN has a bot protection setting that challenges automated visitors, which quietly turns agents away at the door.

From an agent's point of view, your store is a locked shop with the lights off. It can't confirm the price, can't compare the size, can't see the delivery cost or the returns terms, and may not get in at all. When a buyer asks for "a weatherproof eight-seat outdoor dining setting delivered to Fremantle," the agent recommends a competitor whose data it can read.

The fixes are mostly unglamorous. Move specs into text and mark them up. Show delivery estimates on the product page. Turn the returns PDF into a plain page. Add availability to your product schema. Review the bot settings and allow the agents you want. Then check your platform's agentic commerce options and switch on what's available. None of that requires rebuilding the store. All of it makes you readable, reachable and ready.

If you do not sell products online

Most of this still applies, in a different shape. Service businesses are not exempt from agentic discovery, they are just at an earlier stage of it.

The agent is not completing a transaction. It is shortlisting, comparing and often booking. Which means your equivalent of product data is your service data. What you do, where, for whom, at what price range, with what availability and what guarantee.

Publish that in structured, machine readable form and you are doing the same work under a different name.

For a service business, that usually means a clear service page for each offering, a published service area, price ranges or starting prices in plain text where you can, online booking that doesn't hide behind a login, and consistent details across every listing. It also means knowing what the assistants currently say about you, which you can check in minutes using the method in what AI is telling your customers about you.

The three questions to ask your platform

You do not need to become an expert in any of this. You need three answers from whoever runs your website or store.

What structured data are we currently publishing about our products or services, and is it complete and accurate. Which agentic commerce integrations does our platform already support, and are they switched on. And are we accidentally blocking legitimate agent traffic at the firewall or CDN level.

If your developer cannot answer those quickly, that is the project. Everything else in this article can wait until those three are settled, because none of the protocols matter if agents cannot read you or reach you.

The mistakes that cost agentic traffic

• Waiting for a winner. The readable, reachable groundwork pays off under every protocol. Waiting just hands the early position to competitors.
• Betting on a single standard. Every protocol you skip is a group of agents that may not be able to reach you.
• Locking key facts in images and PDFs. Agents compare on specs, delivery and returns. If they can't read them, you lose the comparison by default.
• Leaving security defaults unreviewed. Blocking every bot felt safe. Now it can block buyers.
• Not measuring it. Set up tracking for visits referred by AI tools so you can see whether this work is paying off. We look at what that traffic does once it lands in why AI shoppers bounce and how to keep them.

The timing question

Is this early? Yes. Standards are competing, shapes are changing, and some of these acronyms will not exist in two years.

But notice that everything on the practical list above is work you should do anyway. Clean structured data, readable terms, unblocked crawling and accurate service information all improve conventional search, AI answers and human conversion at the same time.

There is no version of the future where badly structured product data is an advantage. Do the boring layer now and whichever protocol wins, you are already legible to it.

If you'd like a team to do the boring layer properly, that's squarely what our web development work covers: structured data, readable product and service pages, platform integrations and a sensible review of what your firewall lets through. Get readable, get reachable, then let the protocols settle.

Frequently asked questions

What is the Agentic Commerce Protocol (ACP)?

ACP is the protocol behind OpenAI's shopping surfaces in ChatGPT. In its current form, the assistant handles discovery, finding and presenting products, then hands the buyer to the merchant's own site to complete checkout. An earlier instant checkout model was retired in favour of this approach, so merchants keep the transaction while the assistant handles the research.

Do I need a developer to support agentic commerce protocols?

Often not at first. If you're on Shopify or another major platform, much of the support comes through settings and product feeds rather than custom development. The groundwork that matters most is clean structured data, readable terms and making sure agents aren't blocked. A developer becomes useful for custom stores, complex catalogues or reviewing firewall and CDN settings.

Is my website blocking AI agents?

It might be. Many firewalls and CDNs include bot protection that challenges or blocks automated visitors, and robots files sometimes exclude AI crawlers. Ask whoever manages your site to review which bot categories are blocked and what your robots file allows. Then decide deliberately which agents you want to let in, rather than relying on security defaults.

Why should I support more than one agentic commerce protocol?

Because the standards are still competing and nobody knows which will dominate. A merchant supporting only one protocol can be invisible to agents and payment systems built on the others. Spreading your support means you're reachable through more assistants and payment systems, and you're not exposed if one standard fades.

That game is not dead, but it has been demoted. Something else now decides whether a buyer ever hears your name.

When someone asks an assistant who they should hire, the answer is assembled from what other people have said about you. Reviews, forum threads, roundups, comparison articles, directory listings, a comment on a local Facebook group from 2024.

Your website is one voice in that assessment. Everyone else is the rest of the choir.

So yes, reviews now do the job backlinks used to. AI assistants read what customers and third parties say about you before deciding whether to recommend you, which makes detailed reviews spread across several platforms one of the strongest visibility levers a business has.

Which is why a business with a mediocre website and 200 detailed reviews can now beat a business with a beautiful website and eleven.

Why third party sources carry so much weight

Put yourself in the model's position. You have two sources describing the same business. One is the business itself, saying it is the leading provider of quality solutions. The other is forty seven customers describing specific jobs, specific problems and specific outcomes.

Only one of those is useful for making a recommendation, and only one of them carries any risk if it is wrong.

These systems have good reason to be cautious about self description, because self description is where marketing lives. Independent corroboration is the thing that turns a business from a name in the index into a name in the answer.

It's the same logic that made links valuable in the first place. A link was a vote from someone else. A detailed review is a vote with reasons attached, and reasons are exactly what an assistant needs to explain why it's recommending you. That's also why ranking well on Google no longer guarantees you a place in the answer, a shift we unpack in why ranking number one no longer gets you into the answer.

On your website you control the message. In the AI answer, the internet's consensus about you is the message. The only question is whether you have shaped it.

The four sources that actually get read

Reviews with detail. Not star ratings. Text. "Five stars, great job" contributes almost nothing, because it contains no fact. "They rewired a 1970s cottage in Umina in three days, worked around our tenants and came in eighty dollars under quote" contributes an enormous amount. It names the service, the location, the timeframe, the constraint and the outcome.

Roundups and listicles. Third party "best of" articles are catnip for assistants because they are already structured as a ranked recommendation. In Peec AI's analysis of nearly 200,000 AI responses, ranking first in a frequently cited listicle lifted a B2B software brand's visibility in AI answers by 16.5 percentage points, the biggest lift of any category it measured. Somebody else's article is now prime real estate for your brand.

Forums and communities. Reddit threads, trade forums, local groups, question and answer sites. These read as unfiltered peer opinion, and models weight them accordingly.

Industry and supplier mentions. Being listed as an authorised installer, a certified partner, a member of an association. Dull, and unusually persuasive to a machine looking for corroboration.

Together, these sources are how a model decides you're a real, established entity worth naming. If you want the bigger picture on that, read how to make AI trust your brand.

How to engineer detailed reviews instead of star ratings

The default review request produces the least useful review possible. "Please leave us a review" gets you five stars and four words.

Change the ask and the output changes completely.

• Ask a question, not for a favour. "If you have two minutes, what were you worried about before you booked us, and what actually happened." That prompt reliably produces three or four sentences containing real facts.
• Ask within 48 hours of the result. Detail decays fast. A week later you get generic gratitude.
• Make the specifics easy. Remind them what you did. "Following the roof restoration in Terrigal last Thursday" gives them the details to include without you writing the review for them.
• Spread across platforms. Google, industry directories, Facebook, and any marketplace relevant to your category. A model triangulating across four sources is far more confident than one reading a single profile.
• Reply to every one, with substance. Your replies are public too, and a thoughtful reply adds another paragraph of specific, attributable text about your business.

Build the review request into your process

Good intentions produce a burst of reviews and then nothing. A system produces them every week. Here's how to set one up:

1. Pick the trigger moment. The point where the customer is happiest: job finished, product delivered and used, result achieved. Put the review request there, every time.
2. Write the message once. A short text or email using the question-based ask above, with the job details filled in and a direct link to the review page.
3. Automate the send. Most CRMs, booking systems and ecommerce platforms can send it automatically when a job is marked complete.
4. Rotate the platform. Send some customers to Google, some to an industry directory, some to Facebook, so your coverage grows everywhere rather than on one profile.
5. Follow up once. A single polite reminder a few days later. No more.
6. Review the results monthly. Read the new reviews. Are they detailed? Do they mention services and locations? Adjust the prompt if they're thin.

One rule that isn't negotiable: never buy reviews, write them yourself or offer rewards in exchange for positive ones. Google's review policy prohibits it, the ACCC says fake reviews are against the law, and a profile that looks manufactured undermines the very trust you're trying to build.

Getting onto the lists

This part feels uncomfortable to most business owners and it should not. Publishers of roundup articles are actively looking for businesses to include, because a "best of" list with four entries is a bad article.

Find the roundups that already rank for your category and city. Read what the included businesses have in common. Then email the publisher with something genuinely useful, a clear description, real credentials, a photo, a specific differentiator and a customer they can contact.

Do that ten times and you may well land a few. A few third party listings can be a bigger visibility win than a year of blog posts on your own domain, because it is corroboration rather than assertion.

The consistency rule that ties it together

One last thing, and it undoes a lot of good work when ignored.

Every source describing you needs to agree. Same business name, same service area, same core services, same phone number. When a model finds three different service areas across four sources, it has no clean answer to give. That makes it easier to name someone else.

Spend an afternoon auditing every profile, directory and listing you can find with your name on it. Fix the stale ones. Kill the duplicates. This is deeply boring work and it is worth more than most campaigns.

Handling the bad one

You will get a bad review. Everyone does, and the way you handle it is read by both humans and machines.

Reply quickly, without defensiveness, with specifics. Acknowledge what happened, say what you did about it, and offer to resolve it. Never argue the facts in public even when you are right.

A thoughtful reply to a two star review is often more persuasive than the five star ones above it, because it is the only evidence a prospect will ever get about what happens when something goes wrong.

And a profile with a handful of imperfect reviews reads as real. A wall of flawless five stars reads as bought, to buyers and to the systems reading them.

A worked example: the physio with the thin profile

Here's a hypothetical. Say you run a physiotherapy clinic in Newcastle with a strong website, a decent Google ranking, and thirty reviews that mostly say "great service, highly recommend." When you ask ChatGPT for a good physio in your suburb for running injuries, a competitor with a plainer website gets named instead. Their reviews talk about shin splints, marathon training plans and getting back to parkrun.

Over the next quarter, you change the ask. Every patient finishing a treatment plan gets a text asking what they came in with and what they can do now. You send some to Google and some to a health directory. You reply to each review with a line about the treatment approach. You email two local running clubs that publish recommended provider lists, and you fix an old directory listing that still shows your previous address.

Nothing on the website changed. But now there's a growing body of independent text connecting your name to running injuries, your suburb and real outcomes. That's what you'd watch: the specificity of new reviews, the number of platforms carrying them, and whether assistants start naming you when you test the same questions each month.

How to measure whether it's working

• Test the prompts monthly. Ask ChatGPT, Perplexity and Gemini the questions your customers ask. Record whether you appear and how you're described. Our guide to what AI is saying when customers ask about you covers how to do this properly.
• Track review volume and detail across platforms. Not just the star average. The length and specificity matter more.
• Count third party listings. Roundups, directories and partner pages that mention you.
• Watch AI referrals and branded search. More recommendations should show up as more people arriving from assistants or searching your name. We explain how to see that traffic in why traffic from AI converts better.

The reframe

Stop thinking of reviews as reputation management, a defensive chore handled when something goes wrong.

They are now your distribution. They are the raw material from which the machine builds the sentence that decides whether a buyer ever types your name.

Forty detailed reviews is a marketing asset. Four hundred is a moat. Our digital marketing team builds the review systems, listing outreach and consistency audits that grow that asset month after month.

Frequently asked questions

Do reviews help you show up in ChatGPT and other AI answers?

Yes. AI assistants build recommendations from what third parties say about a business, and reviews are one of the richest sources available. Detailed reviews that mention specific services, locations and outcomes give the model concrete reasons to recommend you. Star ratings alone contribute little. Reviews spread across several platforms carry more weight than a large number on a single profile.

How do I get customers to leave more detailed reviews?

Ask a question instead of asking for a favour. Something like "what were you worried about before you booked us, and what actually happened" prompts specific, useful answers. Send the request within 48 hours of the result, remind the customer what you did for them, include a direct link, and follow up once. Automating the request through your booking or CRM system keeps reviews arriving consistently.

Which review platforms matter most for AI search?

Google is usually the most important, but spread matters. Industry directories, Facebook, relevant marketplaces and trade association listings all give AI systems more independent sources to cross-check. A business described consistently across four platforms looks more trustworthy than one with all its reviews in one place. Make sure your name, service area and contact details match everywhere.

Should I reply to negative reviews?

Always, and quickly. Acknowledge the issue, explain what you did about it and offer to resolve it privately, without arguing the facts in public. Prospects and AI systems both read replies. A calm, specific response to a critical review often builds more trust than the positive reviews around it, because it shows how you behave when something goes wrong.

Google is auto migrating older Search setups into AI Max, starting with campaigns using automatically created assets and the campaign-level broad match setting, with Dynamic Search Ads following in February 2027. Microsoft went further and made its own AI Max the default for new Search campaigns.

So should you fight AI Max or accept it? Accept it, on your terms. Fix what the system is optimising toward, feed it your real sales outcomes, and keep hold of budget, negatives and message before you let it run the mechanics. Handled that way it can outperform manual management. Handled carelessly it buys you cheap leads that never close.

You can switch individual features off. You cannot switch off the direction. The platforms have decided that keyword level control belongs to a machine now, and the only real decision left to you is which levers you hold onto while the rest goes automatic.

Get that decision wrong and you will spend six months buying leads you cannot use.

What AI Max actually does to your account

Strip away the branding and it is three changes bundled together.

It broadens matching well past your keywords, using the intent of the query and the content of your landing pages rather than the terms you selected. It writes and assembles headlines and descriptions on the fly. And it decides which URL to send each query to, sometimes overriding the one you specified.

In other words, it takes over the three jobs that used to define search advertising. What you target, what you say, and where you send them.

For some accounts this is a genuine gift. If you sell one thing to a broad market and every enquiry is roughly equal in value, handing over the mechanical work usually improves results and frees up your week.

For others it is a slow disaster, and the pattern is predictable.

The accounts that tend to struggle look like this:

• Lead value varies wildly. A builder whose jobs range from a small repair to a full renovation cannot treat every enquiry as equal. The system can, and will, unless you tell it otherwise.
• The conversion signal is thin or dirty. Duplicate tags, page views counted as conversions, phone clicks counted as calls. The machine learns from whatever you feed it.
• The service area or niche is narrow. Broad matching drifts. If you only serve three suburbs or one specialist product, drift is expensive.
• The landing pages are vague. Because AI Max reads your pages to decide what to match, a page that lists every service you have ever offered invites every search you never wanted.

If you recognise your account in two or more of those, the levers below are not optional.

Automation optimises toward the target you set. If your target is "a form submission" it will find you an infinite supply of people who fill in forms and never buy.

The four levers that must stay yours

One. The conversion definition. This is the whole game and it is where most accounts lose control. If the system is optimising for raw lead volume, it will deliver raw lead volume, cheerfully, at a lower cost per lead than you have ever seen, and your sales team will spend three weeks talking to people who wanted a free quote for something you do not sell.

Feed it qualified events instead. A booked appointment. A quote accepted. A sale, with value attached. Offline conversion imports exist precisely for this and far too few small accounts use them. Do that and the automation becomes an asset, because it is now hunting for the thing you actually want.

Two. Budget and the pace of change. Automated systems learn by spending. That is fine when you have decided how much tuition you are paying. Set the budget deliberately, change one thing at a time, and give each change enough runway to produce a readable result. Constant fiddling resets learning and guarantees you never find out what worked.

Three. Negative terms and brand safety. Broad matching pulls in queries you would never have chosen. Some of those are gold. Some are people looking for free advice, DIY instructions, jobs, or your competitor's warranty department. Review the search terms report weekly for the first two months and build the negative list aggressively. This is the single highest value hour in the account.

Four. The message and the landing experience. The system will generate combinations of your assets. It will not invent your positioning, your guarantee or your offer. Those are yours. Give it strong raw material and it assembles something good. Give it generic material and it assembles something generic, faster.

The migration checklist

If a migration is coming, or has already happened, work through this before you judge the results.

• Fix your conversion tracking first. Nothing else matters if the signal is wrong. Confirm what counts as a conversion, remove duplicates, and stop counting page views and phone clicks as if they were sales.
• Import your real outcomes. Push qualified leads and closed revenue back into the platform. This is the difference between automation working for you and working on you.
• Clean your landing pages. The system now reads them to decide what to match against. A vague page produces vague matching. Make each page unambiguous about what it sells and to whom.
• Set up a search terms review ritual. Weekly, in the diary, for eight weeks. After that, fortnightly.
• Keep one control. Where budget allows, hold a tightly controlled campaign on your highest value terms so you have something honest to compare against.

A worked example: the kitchen renovator

Here is a hypothetical to make the stakes concrete. Say you run a kitchen renovation business in Canberra, spending $6,000 a month on Google Search. Your old campaign was built on tight keywords like "kitchen renovation Canberra" and "custom kitchen quote". A typical job is worth tens of thousands of dollars. Your conversion action is "contact form submitted".

The account gets migrated. Within a few weeks the dashboard looks fantastic. Conversions are up, cost per conversion is down. But your sales team is miserable. The new enquiries include people wanting a cabinet door hinge replaced, DIY flat-pack questions, a few job seekers, and a handful of people in towns you do not service. Every one of them filled in a form, so every one counted as a win.

Now apply the four levers. You change the primary conversion to "site measure booked" and import signed contracts with their value from your CRM. You add negatives for jobs, DIY, flat-pack, hinges and repairs. You rewrite the landing page so it says plainly that you do full kitchen renovations, in the Canberra region, from a stated starting price. You hold the budget steady for a month while it relearns.

Cost per lead goes up. That feels bad for a week. But the leads are now people planning a real renovation, and cost per signed job, the only number that pays your bills, heads the right way. Same automation. Completely different result, because you changed what it was hunting for.

The branded search question

One specific trap worth naming. Broad automated matching loves branded searches, because they convert brilliantly and make the whole campaign look like a triumph.

They were also going to convert without you paying for them. If you are not separating branded from non branded performance, the automation will happily spend an increasing share of your budget buying traffic you already owned, and the dashboard will glow.

Split it out. Judge the two separately. This alone recovers real money in a lot of accounts.

How to read the results without fooling yourself

Automated campaigns come with automated reporting, and that reporting is written by the same system that wants more of your budget. Read it with a raised eyebrow.

• Judge on revenue, not conversions. The metric that matters is cost per sale or return on ad spend from real closed revenue. If you cannot see that yet, fixing it is your first job, and the quality of the signal you send decides how well any automated system performs.
• Compare against your control. The controlled campaign on your highest value terms is your honest baseline. If the automated campaign cannot beat it on cost per sale, it has not earned more budget.
• Watch the search terms, not just the totals. Totals can look healthy while the mix quietly shifts toward cheaper, worse queries.
• Give it time, then be decisive. A few weeks is not enough to judge a learning system. A few months of flat or falling sales quality is enough to pull back.

If you suspect a big share of the budget is already being burned, run the full audit in where 30% of ad spend goes. It pairs neatly with the migration checklist above.

What to do this week

1. Log in and check every campaign type. Find out which campaigns have been migrated, or are scheduled to be.
2. Open your conversion settings. Write down exactly what each campaign is optimising toward. If it is form fills or page views, flag it.
3. Pull last month's search terms report. Highlight every query you would never have chosen. Add the obvious ones as negatives today.
4. Split branded from non branded. Look at the non branded numbers on their own. That is your real performance.
5. Book the diary slot. One hour every week for eight weeks, for search terms and results. Put it in now or it will not happen.
6. Plan the offline import. Work out how qualified leads and closed sales will flow from your CRM back into the platform, even if it takes a few weeks to set up.

The honest verdict

Automated search is not a scam and resisting it entirely is not a strategy. On clean signal, with a decent offer and strong landing pages, it can outperform plenty of manual account management. The machine is better than you at the mechanical parts, and it never gets tired at 4pm on a Friday.

What it cannot do is decide what a good customer looks like, what you are willing to pay for one, or what makes you worth choosing.

Hand over the mechanics. Keep the judgement. That division of labour is the entire skill of running paid media from here on, and the accounts that get it right will quietly outperform the ones that either fought the change or surrendered completely.

Go and look at what your account is optimising toward today. If the answer is "form fills," you have found your afternoon.

And if the ads are fine but the money still leaks, look at the other end of the click. The system is only as good as the page it sends people to, which is why a landing page built to convert matters more under automation, not less. If you would rather have someone set the signal, the guardrails and the review rhythm for you, that is exactly the kind of advertising work LIQID NOISE does: hand the mechanics to the machine, keep the judgement with people who care what a good customer looks like.

Frequently asked questions

What is Google AI Max for Search campaigns?

AI Max is a set of automation features Google applies to Search campaigns. It broadens matching beyond your chosen keywords using the query's intent and your landing page content, generates headlines and descriptions on the fly, and can choose which of your URLs to send each searcher to. In short, it automates targeting, ad copy and landing page selection.

Can I opt out of AI Max in Google Ads?

For now, Google lets you toggle AI Max features off in campaign and ad group settings, but the direction is clear: platforms are moving search advertising toward automation. Rather than spending energy resisting, focus on controlling the inputs that matter. Define conversions around qualified leads and sales, import offline outcomes, build strong negative keyword lists, and keep one tightly controlled campaign as a benchmark.

Why did my leads get worse after switching to AI Max?

Usually because the system is optimising toward the wrong goal. If a form submission counts as a conversion, it will find the cheapest people who fill in forms, whether or not they ever buy. Switch the goal to qualified events such as booked appointments or sales with value, import those outcomes, and review search terms weekly.

How often should I check the search terms report with AI Max?

Weekly for the first eight weeks after a migration or new launch, then fortnightly once the negative list is solid. Broad matching pulls in queries you never chose, and some will be expensive and irrelevant. A regular review is one of the highest value hours you can spend in the account.

That website is now being read by something else entirely, and it hates almost everything on it.

AI assistants do not rank pages. They extract statements. They pull sentences out, check them against other sources, and reassemble them into an answer with a name attached. If your page contains no extractable statement, you are not competing badly. You are not in the pool.

The best known research on what gets extracted points the same way, and it comes down to three edits: put real numbers on the page, quote named humans, and cite where your facts come from.

In the GEO study from Princeton and partners, published at KDD 2024, adding quotations lifted a page's visibility on the researchers' test engine by around 41%. Adding statistics lifted it around 33%. Citing sources lifted it around 28%.

Almost no business website does any of the three.

Edit one. Put a number in it.

Open your main service page. Count the numbers. Not the phone number and not the year you were founded. Numbers that describe what you do or what happens when you do it.

Most pages have zero. They have adjectives instead. Fast, reliable, affordable, experienced, trusted.

Adjectives are unquotable. A model cannot repeat "we are reliable" as a fact, because it is not one. It can repeat "we respond to commercial callouts within four business hours across the Central Coast" all day long, because that is a claim with an edge on it.

Go through your key pages and convert every adjective into a measurement.

• "Fast turnaround" becomes "most jobs completed within five working days."
• "Affordable" becomes "typical projects run between four and nine thousand dollars."
• "Experienced" becomes "over 600 installations since 2014."
• "Trusted" becomes "142 Google reviews at 4.9 stars."

Every one of those is a sentence an assistant can lift verbatim to justify recommending you. Every adjective you replace is one less reason for the model to reach for a competitor who was more specific.

One rule while you do it: the number has to be true, and you have to be able to stand behind it. If you do not know your average turnaround, go and find out. Pull it from your job system, your invoices or your calendar. An invented number is worse than an adjective, because a model may repeat it to a buyer, and the buyer will hold you to it. Specific and honest is the target. Specific and made up is a liability waiting for a phone call.

A model cannot quote a feeling. It can only quote a fact. Every vague sentence on your site is a sentence that will never be read aloud to a buyer.

Edit two. Quote a human, by name.

Quotation marks do something specific to a language model. They mark a passage as attributable, self contained and safe to reproduce. That is why quoted content gets pulled through at a materially higher rate.

You do not need to interview an industry analyst. Three sources are sitting in your business right now.

Yourself. A short, opinionated, named quote in the middle of a page. "We stopped offering the cheap option in 2023 because it was generating 80% of our warranty callbacks, says Kendall King, founder." That is quotable, attributable and human.

Your customers. Real testimonials with a full name and a specific outcome, published as text on the page rather than trapped inside an image or a third party widget the model cannot read.

Your team. The person who actually does the work usually has the most quotable sentence in the building, because they talk about the job rather than the marketing of the job.

Format matters as much as source. Put the quote in the page text, in quotation marks, with the full name and role straight after it. Keep it short enough to stand alone and specific enough to be worth repeating. "Great service, highly recommend" is a quote in the technical sense and useless in every other sense. "They rewired our kitchen in a day and left it cleaner than they found it" is something a model can actually use to describe you.

Edit three. Cite where things came from.

This is the one businesses resist hardest, because linking out feels like sending people away. That instinct is a decade out of date.

Citation does two things. It signals to a model that the page is assembled from verifiable material rather than invented, and it places your page inside a network of sources the model already trusts.

Practically, it means when you state an industry figure, say where it came from and link it. When you refer to a standard, a regulation or a warranty term, name the document. When you make a claim about typical results, say what it is based on.

A page that reads like a well researched briefing gets treated as a source. A page that reads like a brochure gets treated as marketing, and marketing is exactly what these systems are built to filter out.

The structure that makes extraction easy

Beyond the three edits, formatting decides whether a model can cleanly lift what you wrote.

Lead every page and section with a direct answer in the first two sentences. Not context. Not a warm up. The answer, then the explanation.

Use headings that are literally the questions people ask, then answer them immediately underneath. A model matching a user's question to a heading it recognises is doing pattern matching, and you can just hand it the pattern.

Keep statements self contained. "As mentioned above, this applies here too" is worthless to an extraction system, because the sentence means nothing on its own. Repeat the subject. It reads slightly redundant to a human and perfectly to a machine.

And define your terms plainly, once, near the top. A clean definition is exactly the kind of self contained passage an extraction system can lift.

A worked example: the accountant's services page

Say you run a hypothetical accounting practice in Adelaide that works with tradies. Your main services page says you are "experienced, affordable and focused on helping small businesses thrive." Ask an assistant who helps tradies with tax in Adelaide and you are not mentioned.

Now apply the three edits. The adjectives become measurements you can verify from your own records: how many trade businesses you currently look after, your typical turnaround for a quarterly activity statement, and a fixed fee range for a standard sole trader return. A named quote goes in from you as the principal, explaining the single mistake you see tradies make most often with vehicle deductions and why it costs them. A second quote comes from a real client, by name, with a specific outcome in their own words. Then you cite properly: when you mention the instant asset write-off or record-keeping obligations, you name and link the relevant government guidance rather than paraphrasing it from memory.

Finally, the structure. The page opens with a two-sentence answer to "what does a tradie accountant do and what does it cost?" The headings become the questions clients actually ask. Each section's first line answers its heading directly. Nothing about the practice changed. What changed is that the page now contains a dozen statements a model can lift, check and attach to your name.

A warning about overcorrecting

One caution, because this advice gets taken too far.

Writing for extraction does not mean writing like a database. Pages stuffed with disconnected facts and no argument read badly to humans, and humans are still the ones who decide to pay you.

The goal is a page that works twice. A person reads a clear, persuasive argument. A machine finds a dozen clean, quotable facts inside it.

The good news is these rarely conflict. Specific writing with real numbers and named sources is simply better writing. The extraction benefit is a side effect of doing the thing you should have been doing anyway.

Do this on five pages, not fifty

You do not need to rewrite the site. Pick the five pages tied to actual revenue.

For each, spend an hour. Replace every adjective with a number. Add one named quote. Cite one external source properly. Rewrite the opening two sentences so they answer the question in the title. Convert the headings into the questions your customers actually ask.

Five hours of work, on the five pages that matter, against a system that is currently deciding what to tell buyers about your industry.

Your competitors are still adding keywords to their meta descriptions. Let them.

How to tell whether it is working

Before you edit, write down five questions a buyer would ask an assistant about your service and your area. Ask them in ChatGPT, Perplexity, Gemini and Google's AI answers. Note who gets named and which sources get cited. Make your edits, then ask the same questions again every month.

Watch for three shifts: your business starting to appear in answers, the description of you becoming more specific and accurate, and your own pages showing up in the cited sources. Also check your analytics for referral traffic from AI assistants, which tends to arrive with a clearer intent than a cold search click. Changes to assistants that search the live web can show up relatively quickly. Changes that depend on training data take longer, which is one more reason to start now. For why ranking and being cited no longer move together, see why ranking number one no longer gets you into the answer, and for the wider strategy behind these edits, read the GEO playbook for getting cited by ChatGPT, Gemini and Perplexity.

Make your best pages the ones machines quote

None of this requires a new website. It requires a sharper one: pages that say specific, true, attributable things in a structure a machine can lift. That is simply good writing with the vague parts removed. If you want more on the craft of quotable writing itself, write content AI can't help but quote goes further. If you would rather hand it over, our digital marketing team rewrites revenue pages for humans and extraction at the same time, then tracks what the assistants start saying about you.

Frequently asked questions

How do I get my business quoted by ChatGPT?

Give it statements worth quoting. Replace vague adjectives on your key pages with specific, verifiable numbers, add short named quotes from you, your team and real customers, and cite the sources behind any industry claims. Open each page with a direct answer and use headings that match real customer questions. Then check monthly what assistants say about your business and category.

Does adding statistics to a page really help AI visibility?

Yes. The GEO study published at KDD 2024 found adding statistics lifted visibility on the researchers' test engine by around 33%, alongside gains from cited sources and quotations. The reason is simple: a number is a self-contained, checkable fact a model can repeat with confidence. Only use statistics you can verify, ideally from your own records, and state what each number is based on.

Should I rewrite my whole website for AI search?

No. Start with the five pages most tied to revenue, usually your main service pages and the page that converts best. An hour per page applying the three edits and fixing the structure is enough to test the approach. Measure what assistants say before and after, then roll the method out to more pages once you can see it working.

Will writing for AI make my website worse for customers?

Not if you do it properly. Specific numbers, named quotes and cited sources make pages more persuasive to humans too. The risk is overcorrecting into a page of disconnected facts with no argument. Aim for a page that reads as a clear, confident case to a person while containing plenty of clean, quotable statements for a machine.

It is harder than it was, and it is getting harder monthly. Perfect skin, impossible lighting, hands that are nearly right, a kitchen that does not exist in any country. The whole internet has developed a faint plastic sheen.

Most businesses looked at that and thought about how much money they would save.

The smart ones looked at it and saw the opening. When everything looks generated, proof that a human was actually there becomes the scarcest thing in the feed.

That scarcity has a price. Call it the authenticity premium, and it is the cheapest competitive advantage available to a small business right now.

Put plainly, the authenticity premium is the extra trust, attention and willingness to pay that buyers now give to visuals they believe are real. You earn it by showing your actual work, your actual people and your actual place, consistently, in ways a generator cannot fake.

What actually changed in the buyer's head

Trust in imagery used to be automatic. A photo of a finished job was evidence a job had been finished. That inference is broken now, and buyers have quietly recalibrated.

The recalibration goes further than most marketers realise. It is not just that people distrust obviously generated images. It is that they now distrust polish itself, because polish is the thing generation is best at.

So the glossy hero shot that signalled professionalism in 2020 now signals something closer to "we typed a prompt." Meanwhile the slightly crooked photo taken on site, with a cloudy sky and a ute in the background, reads as real.

Think about how you behave as a buyer now. You see a perfect photo of a renovated bathroom and your first thought is no longer "nice work." It is "is that even theirs?" You go looking for something that settles the question. A video walkthrough. A review with a photo attached. The same tiler appearing across several jobs. Your customers are doing exactly the same thing to you, and if your visuals cannot answer the question, they quietly move on to someone whose visuals can.

The market has not stopped valuing quality. It has stopped accepting polish as proof of it. Those are very different things and the gap between them is where you win.

The five signals that read as real

This is not an argument for lazy photography. It is an argument for a specific kind of deliberate imperfection. Five things consistently signal that a human was present.

• Context that could not be invented. The specific street, the recognisable local landmark, the weather that day. Generic backgrounds read as generated because generated backgrounds are generic.
• Continuity across shots. The same person, the same van, the same worn tool showing up across a body of work. Generation still struggles to hold one consistent world across a whole body of work. A real business does it without trying.
• Faces you see repeatedly. Not a model. Your team. Recognition builds over months and is one of the hardest things to fake.
• Motion and sound. Handheld video with real ambient audio is still more work to fake convincingly than a single still. It is also cheaper to produce than it has ever been.
• Visible process, not just results. The half finished stage, the mess, the problem being solved. Anyone can generate a beautiful outcome. Almost nobody bothers to fabricate the awkward middle.

Where AI imagery still earns its keep

To be clear, this is not a purity argument. Generated imagery is genuinely useful in places where nobody is being asked to believe anything.

Backgrounds, textures, abstract graphics, concept exploration, mood boards, internal decks, quick mock ups to show a client an idea before committing to a shoot. All fine. Nobody is deceived and nobody cares.

The line is simple and worth writing down. Anywhere a customer might reasonably take the image as evidence, it has to be real. Product shots, results, team, premises, testimonials, before and after. Get caught faking one of those and you do not lose an image. You lose the credibility of every image you have ever published.

That risk is asymmetric and it is not worth the saving.

We went deeper on the specific dangers, from platform labelling to regulators, in why using AI product photos could torch your brand. The short version: use the generator for the scaffolding, never for the evidence.

Building an authentic library on a small budget

The objection is always cost. It should not be, because the cost dropped just as fast as the generated alternative.

A modern phone shoots footage that would once have needed dedicated camera gear. What you need is a system, not equipment.

Here is the system, step by step.

1. Write the shot list from your sales conversations. List the five questions or doubts prospects raise most. Each one becomes a shot. "Do you clean up after?" becomes a clip of your team leaving a site spotless.
2. Pick one person to own it. Not a committee. One person who is responsible for making sure the photos get taken and filed.
3. Set a simple standard. Clean lens, shoot in daylight where possible, hold the phone steady, capture both vertical and horizontal.
4. Get permission. Ask customers before you photograph them or their property, and keep a record that they agreed.
5. File it the same day. A shared folder sorted by month and job type. Unfiled photos are photos you will never use.

Run a shoot day once a quarter. Half a day, real jobs or real customers, a shot list built from your actual sales objections rather than a mood board. Capture stills, short vertical clips, ambient audio and at least one unscripted customer answer.

Between shoot days, make photos part of the job. Every completed project gets three frames, before, during and after, taken by whoever is on site. Twelve months of that discipline produces a library no competitor can replicate, because it is a record of work that actually happened.

Then use it everywhere. The reason most businesses fall back on stock is not that they lack images. It is that the images live in a phone gallery nobody has ever sorted.

One good shoot day can feed your channels for months if you plan it that way, which is exactly the approach in one video shoot, 90 days of content. And do not overthink the production value. As we have argued before, "ugly" phone videos often outsell the polished commercial, precisely because they look like they happened.

A worked example: the landscaper's library

Say you run a landscaping business on the Sunshine Coast. Your website uses a handful of beautiful stock garden photos and one AI-generated hero image of a backyard at sunset. It looks lovely. It also looks exactly like every other landscaper's site in the region.

You start small. Every job gets a before, a during and an after, taken on the crew leader's phone. Once a quarter you spend half a day filming two finished gardens with the owners, asking them one unscripted question each: what made you pick us? You capture the sound of the site, the crew talking, the rain on the new paving.

After a year you have a library of real gardens, on real streets, built by the same recognisable crew, with homeowners in their own words. You replace the stock images on your homepage with that work. Your social feed stops being quotes and stock photos and becomes a running record of jobs. Nothing about it is expensive. All of it is impossible for a competitor to copy, because they did not do those jobs.

The compounding advantage

Here is why this matters more each quarter rather than less.

Generated imagery gets cheaper and more available every month, which means the supply of polished content is heading toward infinite. Anything infinite has a price of zero.

Meanwhile the supply of genuine documentation of your actual work is fixed. It can only be produced by doing the work and being there with a camera. That makes it the only visual asset in your business that appreciates.

Five years of consistently documented real work is a moat. Five years of stock photography is a subscription.

Where this is heading

Expect the gap to widen rather than close.

Platforms are adding provenance labelling, with TikTok auto-labelling AI content that carries Content Credentials. Buyers are growing more sceptical of what they see. And as generated content becomes the default, the signals of real production, ambient audio, imperfect light, recognisable places, will read as increasingly deliberate and increasingly premium.

The businesses documenting their real work now are not just avoiding a trust problem. They are quietly building an archive that becomes more differentiating every year, while their competitors accumulate a folder of images that anyone could have made in an afternoon.

How to know it is working

You will not get a neat dashboard for trust, but you will see it in the right places. Watch which posts get saved, shared and commented on, not just liked. Compare enquiry rates on pages that use real work against pages that still use stock. Listen on sales calls for prospects mentioning a specific job, person or video they saw. And ask every new customer what convinced them. When the answers start naming your real work, the premium is paying out.

The test

Open your website. Look at every image on the front page and ask one question of each. Could a competitor have produced this without doing anything I have done.

If the answer is yes for more than half of them, your visuals are not proving anything. They are decoration, and decoration is precisely the thing that just became free.

Go and take some photos of the real thing. It is the strangest competitive advantage of 2026 and it costs almost nothing.

If you want help building the system, or you want a proper shoot day planned around your actual sales objections, that is what our photo and video team does. We capture real work, real people and real places, then turn it into a library that keeps paying you back. Either way, start this week. Every job you finish without a camera out is proof you will never get back.

Frequently asked questions

What is the authenticity premium in marketing?

It is the extra trust and value buyers give to content they believe is genuinely real, now that AI-generated imagery is everywhere. As polished images become cheap and common, proof that a human was present, such as real locations, recognisable staff, visible work in progress and unscripted customers, becomes scarce. Scarce things are worth more, so real visuals now carry a premium.

Is it okay to use AI-generated images on my website?

Yes, for decorative or conceptual uses where nobody is being asked to believe anything, like backgrounds, textures, abstract graphics or mock ups. No, anywhere a customer could reasonably treat the image as evidence, such as products, results, your team, your premises, testimonials or before and after shots. Faking those risks the credibility of everything else you publish.

How can a small business get authentic photos without a big budget?

Use the phone you already have and build a habit rather than buying gear. Photograph every job before, during and after, and run a half-day shoot each quarter focused on real customers and the questions prospects ask most. File everything in a shared folder the same day. Over a year, that produces a library no competitor can copy.

How do customers tell if a photo is AI-generated?

They look for tells like hands that are nearly right, waxy skin, impossible lighting, generic backgrounds and text that melts into nonsense. More often they simply sense something is off, then look for other evidence. Consistent faces, real places, visible process and handheld video with natural sound are the signals that reassure them the business is real.

Then their customer posted an eleven second clip of the product on her kitchen bench, filmed vertically, badly lit, with her kids yelling in the background.

That clip sold eleven units.

This is not an anecdote about luck. In Stackla's survey of 2,042 shoppers across the US, UK and Australia, 79% said content from other customers highly impacts their purchasing decisions. Only 9% said the same of influencer content.

So, do real customers outsell influencers? For most businesses selling to ordinary buyers, yes. Customer videos answer the question buyers actually care about, "will this work for someone like me," while paid creator posts mostly answer "is this cool." Use customers for conversion and creators for reach, and judge each on the job it is suited to.

The reason is not mysterious, and once you see it you cannot unsee it.

Your prospect is not looking for aspiration. They are looking for evidence.

An influencer video answers the question "is this cool." A customer video answers the question "will this work for someone like me."

Only one of those questions is standing between your prospect and their credit card.

When a buyer sees a professional creator in a beautiful kitchen holding your product, their brain files it correctly as advertising. When they see a woman who looks like their neighbour, in a kitchen that looks like theirs, saying the thing they were privately worried about, their brain files it as information.

Polish signals budget. Roughness signals honesty. In 2026 your prospect has been lied to by beautiful footage so many times that beautiful footage is now a warning sign.

This gap has widened in the last eighteen months for one specific reason. Anyone can now generate a flawless studio image of anything for free. So flawless studio imagery no longer proves the product exists, let alone that it works.

That is why a shaky phone clip now carries more weight than a glossy commercial. It is hard to fake a real person in a real room with a real problem. We explored the same shift from the production side in why ugly phone videos outsell your polished commercial, and from the brand side in everything looks like AI now, that is your opening.

What makes customer content convert

Not all of it works. The clips that sell share four traits and it is worth being deliberate about all four.

A specific objection, named out loud. The best customer clips begin with the doubt. "I honestly thought this would be too small for a family of five." That opening does more work than any feature list, because it is the exact sentence running through the viewer's head.

The product in an ordinary environment. Real bench, real driveway, real mess. The setting is the proof.

One concrete result, with a number if possible. "Cut my Sunday prep from two hours to forty minutes." Vague enthusiasm converts nobody.

Their own words. The moment you hand a customer a script, you produce an ad. The value was never the person. It was the unscripted phrasing that no marketer would have written.

If a clip has all four, it will usually beat anything you could have scripted. If it has none, no amount of editing will save it. Before you spend money behind any piece of customer footage, run it past those four traits like a checklist.

How to get it, reliably, without begging

Most businesses treat customer content as a happy accident. It is a process, and it takes about twenty minutes a week.

• Ask at the peak, not at the end. The moment to ask is the moment the customer is delighted, which is usually the day the result lands, not thirty days later when you send the review request.
• Ask for one question, not a video. "Video testimonial" terrifies people. "Can you film a quick answer to one question on your phone" does not. Send the question by text.
• Give them the objection to answer. "What were you worried about before you booked us, and what actually happened." That single prompt produces usable footage far more often than a generic request does.
• Make it worth their while. A discount, a small gift card, a free add on. Paying customers a modest amount for footage is dramatically cheaper than a creator fee and produces better performing assets.
• Get usage rights in writing. One short paragraph by email. You want to run it as a paid ad, on your website, for at least twelve months. Do this every time and you build a library instead of a folder of things you are nervous about using.

The part that separates it from a nice testimonial page

Collecting the footage is half the job. The half that makes money is treating it as paid media raw material rather than social proof decoration.

Take one clip. Cut four versions with four different opening hooks. Run all four. One of them will outperform the others by a distance you will not believe until you see it, and it will rarely be the one you predicted.

A good hook test for customer footage looks like this. Version one opens on the objection ("I thought this would be too small"). Version two opens on the result ("forty minutes instead of two hours"). Version three opens on a caption that calls out the viewer ("for families of five or more"). Version four opens on the most visually surprising moment in the clip. Same footage, same body, four different first seconds. For more on structuring the rest of the ad, see the hook, story, offer formula.

Then put the winner everywhere. Top of the landing page. In the email sequence. In the follow up you send after a quote. On the thank you page. Customer footage is one of the few assets that can improve conversion at every stage of the funnel, because at every stage the prospect is asking the same question. Does this work for someone like me.

How to measure whether it is working

Views are the wrong scoreboard, which is exactly how the four and a half thousand dollar post looked like a success on day one. Judge customer content the same way you judge any paid media.

• Cost per purchase or cost per qualified lead on the paid versions, compared against your existing branded ads over the same period.
• Hold rate past the first three seconds. If people stop but do not stay, the hook is working and the middle is not, or the other way around.
• Landing page conversion rate before and after you add the winning clip above the fold.
• Close rate on quotes where you included a customer clip in the follow up versus where you did not.

Run the comparison honestly. If a branded ad keeps beating the customer footage in your account, keep the branded ad. The point is not ideology. The point is money.

A worked example: the café equipment supplier

Say you sell commercial coffee grinders online to small cafés and spend $3,000 a month on Meta ads. Your ads are polished product renders with a price and a spec sheet. Sales are steady but expensive.

You text your ten happiest customers one question: "What were you worried about before you bought, and what actually happened?" Four reply with clips. One café owner films behind her counter during a rush and says she was worried it would be too loud for a small shop, then grinds a dose while the camera rolls. Another says he nearly bought the cheaper model and explains the difference in his own words.

You cut four hooks from each of the two best clips and run them against your existing render ads with a slice of the budget. You judge only on cost per purchase. The winner goes on the product page, in the abandoned cart email and in the follow up after every wholesale quote. (This business and its numbers are hypothetical, but the process is exactly the one described above.)

When influencers still make sense

To be fair about it, creators are not useless. They are just usually the wrong tool for the job people hire them for.

Creators are excellent for reach into an audience you cannot buy, for borrowing credibility in a new category, and for producing a volume of formats quickly when you have no customer base yet.

They are poor at the bottom of the funnel, which is exactly where most small businesses deploy them and then judge them.

The strongest version of this in 2026 is the hybrid. Long term relationships with a small number of creators who genuinely use the product, producing content in the same rough, native register as your customers, with full usage rights so it can run as paid media. Not a one off post that disappears in a day.

The mistakes that waste good customer footage

• Over-editing it. Colour grading, logo stings and stock music turn a real clip back into an ad. Captions and a tidy trim are usually enough.
• Hiding it on a testimonials page. Almost nobody visits it. Put the proof where the decision happens.
• Running one version. The hook test is where most of the value is. Skip it and you are guessing.
• Forgetting disclosure. If you paid or rewarded someone for their footage, be upfront about it where your local advertising rules require it. Hiding it would undo the very honesty you are selling.
• Letting the library go stale. Ask every month, not once a year. Fresh clips keep your ads from wearing out.

Customer content also does quiet work beyond ads. Reviews and real customer stories are increasingly what AI assistants draw on when they decide who to recommend, which we cover in AI reads your reviews before it recommends you.

Start here

Go through your last twenty customers. Pick the five who were happiest. Text them one question today.

By next week you will have two or three clips. Cut four hooks from each. Put fifty dollars a day behind them.

That is the whole strategy, and it will outperform the four and a half thousand dollar influencer post by a margin that makes the conversation with your accountant genuinely enjoyable.

If you want help turning customer footage into a system, from the ask to the edit to the ad account, that is exactly what our digital marketing team builds at LIQID NOISE. But you do not need us to send the first text. Send it today.

Frequently asked questions

What is UGC in marketing?

UGC, or user generated content, is content made by your actual customers rather than by your brand or a paid creator. It includes phone videos, photos, reviews and social posts showing your product in real use. Businesses use it as social proof and as raw material for ads, because buyers tend to trust people like themselves more than they trust a brand talking about itself.

How do I ask customers for video testimonials?

Ask at the moment they are happiest, usually the day the result lands. Send one simple question by text instead of asking for a "video testimonial," for example "What were you worried about before you bought, and what actually happened?" Offer a small thank you, keep it short, and get written permission to use the clip in ads and on your website.

Do I need permission to use customer videos in ads?

Yes. Even if a customer posted the clip publicly or sent it to you directly, get clear written permission before you run it as an ad or put it on your website. A short email that states where you will use it and for how long is enough for most small businesses. Doing this every time builds a library you can use with confidence.

Are micro-influencers better than big influencers?

Often, for small businesses, because the content tends to feel closer to a real customer and the relationship can be ongoing. What matters most is not follower count but whether the creator genuinely uses the product, produces rough native content and gives you full usage rights to run it as paid media. Judge any creator on cost per sale, not views.

Day twelve it wobbled. Day sixteen the cost doubled. By day twenty you were staring at the account wondering what you broke.

You broke nothing. Plenty of advertisers now see winners tire within two or three weeks, where six weeks or more used to be normal. Your ad did not stop working. It finished working.

So how often should you refresh ad creative now? Plan for new creative every week and expect a winner to carry spend for two to three weeks. That means a production rhythm, not a one-off shoot.

That single change breaks the way most businesses plan, budget and produce marketing, and almost nobody has adjusted for it.

Why the window collapsed

Three forces, all pushing the same direction.

Delivery got faster and smarter. Modern ranking systems find the people most likely to respond to a given creative within days rather than weeks. Once that pocket is saturated, the system has to reach further into colder audiences and your cost climbs. The efficiency you enjoyed was the system spending your budget on the easy wins first.

Feeds got denser. Your prospect's feed is more crowded with ads than ever, because producing an ad now costs almost nothing. Novelty decays faster when everything is competing for the same attention.

And audiences overlap more than ever. Broad targeting is the default now, so your ads hit the same people more often, more quickly.

Fatigue is not the algorithm punishing you. It is the algorithm finishing the job you gave it and asking for more work.

Put those together and the maths is simple. The system finds your best responders faster, shows them your ad more often, and then runs out of easy people sooner. The same thing that made the ad cheap in week one makes it expensive in week three. We explain how Meta's ranking changes feed into this in Meta's Andromeda update.

Telling fatigue apart from three things that look like it

Before you replace creative, make sure fatigue is actually what happened. Four different problems produce a rising cost per result and each needs a different fix.

• Real fatigue. Frequency climbing, click through rate falling, conversion rate from click roughly stable. The people seeing it have seen it. Replace the creative.
• Audience exhaustion. Frequency high across every ad in the account, not just one. Your addressable pool is too small for the budget. Widen the targeting or lower the spend.
• Offer decay. Click through rate holding, conversion rate on the landing page dropping. People still want the click. They stopped wanting the thing. Fix the offer, not the ad.
• Tracking breakage. Results appear to fall off a cliff overnight rather than degrading. Check your pixel, your conversions API and your event match quality before you touch anything creative.

A lot of "my ads stopped working" conversations turn out to be one of the last three. Replacing creative on an offer problem burns money and teaches you nothing.

The tracking one deserves special attention because it is the most expensive to misdiagnose. If the platform is receiving weaker conversion data, it bids and delivers worse, and every creative looks tired at once. Our guide to event match quality shows you where to check.

A quick diagnostic routine

1. Look at the shape of the drop. Gradual decline over days points to fatigue or exhaustion. A cliff overnight points to tracking or a platform change.
2. Check frequency on the ad and on the account. One ad high means that ad is tired. Everything high means the pool is too small.
3. Split the funnel. Compare click through rate with landing page conversion rate. Which one moved tells you whether the ad or the offer is the problem.
4. Confirm the data. Make sure conversions are still being recorded properly before you act on any of it.

The cadence that keeps up

If winners last two to three weeks, your production rhythm needs to be weekly. Not quarterly. Not "when we get around to a shoot."

That sounds expensive until you build it properly, at which point it becomes cheaper than what most businesses currently do.

Shoot in batches, ship in weeks. One structured half day of filming produces enough raw material for six to eight weeks of releases. The mistake is publishing it all at once, which floods the account with near identical creative that the ranking system treats as a single entity anyway.

Run a rolling three. At any time you should have proven creative carrying the spend, fresh creative in test, and the next batch in edit. When the proven one fatigues, the tested one is already warm.

Refresh at the trigger, not the calendar. Set a rule. When frequency passes your threshold and click through drops 20% off its peak, the replacement goes live. Rules stop you from panicking on day six and stop you from clinging on day twenty four.

For the practical side of turning one shoot day into weeks of releases, see one video shoot, 90 days of content.

What to actually change, not just recolour

Here is where most refresh attempts fail. A new colour grade, a different font, the same footage recut is not a new ad. Modern systems cluster near identical creative together, so twelve variants of one idea get treated as roughly one idea.

Change something structural or do not bother.

Change the hook. The first three seconds and the first line of copy carry much of the performance. A price led hook, a problem led hook and a myth busting hook are three genuinely different ads even with identical footage after the fourth second.

Change the format. Static, talking head, screen recording, customer selfie, before and after. Each one reaches a different mood of scrolling.

Change the person. A customer in their driveway and a founder in a studio are not the same ad, even reading the same script.

Change the angle. Speed versus price versus risk reversal versus proof versus the cost of getting it wrong. This is the one that requires actual thinking, and it is the one that produces the outliers.

A worked example: one half day, eight weeks of ads

Say you run a Sydney solar installation company spending $4,000 a month on Meta ads. You have been running the same two videos for three months. They were great in the first fortnight. Now your cost per lead has crept up and nobody on the team can say why.

You book a half day. Before it, you write down five angles: the power bill shock, the installer who never came back (risk reversal), the real customer walking through their system, the myths about panels in winter, and the true cost of waiting another year. For each angle you plan three hooks. On the day you film the founder, one installer and two customers who have agreed to appear. You capture talking heads, a quick screen recording of a monitoring app, before and after roof shots and a few selfie-style clips from the customers.

Back in the edit, that becomes a library of genuinely different ads. You release two or three a week into a testing campaign, promote winners into the main campaign, and replace them when the frequency and click through rule fires. By the time the library runs low, you have learned which angles carry the account, so the next shoot is sharper.

This is a hypothetical, but notice what changed. The budget did not go up. The number of ideas did.

How to measure creative fatigue properly

Set up a simple view in your ads manager you can check in two minutes. For each active ad: days live, frequency, click through rate against its own peak, cost per result against its own best week, and landing page conversion rate.

Watch trends, not single days. A bad Tuesday is noise. Three days of falling click through with rising frequency is a signal. Log every ad with its hook, format, person and angle so that, over time, you can see which structural choices produce winners and how long each type tends to last in your account.

That log is the real asset. It turns creative from guesswork into a system that gets better every month.

Mistakes that make fatigue worse

• Pausing a winner before the replacement is proven. Test the new ad alongside the old one, then shift spend.
• Doubling the budget on a tired ad. More spend on a saturated creative only raises frequency faster.
• Launching ten variants at once. The system splits spend thinly and you learn nothing clear.
• Judging new creative in two days. Give each test enough spend and time to show a real pattern.

The upside nobody mentions

Short fatigue windows sound like bad news. For a small business they are quietly the best news in years.

When winners lasted for months, the advantage went to whoever could afford the big production. You could not out spend them, and the incumbent's polished campaign sat there earning for months.

Now that same campaign can burn out in a few weeks. The advantage moved from production budget to production speed, and a business with a phone, a real customer and something honest to say can move faster than any agency retainer with a two week approval cycle.

Fatigue punishes the slow and rewards the prolific. That used to be the other way around.

That is the kind of creative engine we build in our advertising work at LIQID NOISE: angles planned upfront, batch shoots, weekly releases, clear refresh rules and reporting on what actually made the winners win.

The one change to make today

Open your account. Find your best performing ad. Check the date it launched.

If it is older than three weeks and still carrying your spend, you are not winning. You are running on borrowed time and the replacement has not been made yet.

Book the shoot day. This week.

Frequently asked questions

How long does a Facebook or Instagram ad last before it fatigues?

It varies by account, budget and audience size, but many advertisers now see creative tire within two or three weeks, where six weeks or more used to be common. Bigger budgets and smaller audiences burn through faster. Instead of guessing, set a rule based on rising frequency and falling click through rate, and have replacement creative ready before you need it.

What are the signs of ad creative fatigue?

The classic pattern is frequency climbing while click through rate falls, and conversion rate after the click stays roughly the same. Cost per result rises gradually over several days. If every ad in the account shows high frequency, the problem is more likely audience size. If results drop overnight, check your tracking before blaming the creative.

Is changing the colour or text of an ad enough to fix fatigue?

Usually not. Modern ad systems group near identical creative together, so a recolour or new font often gets treated as the same ad. Change something structural instead: the hook, the format, the person on camera or the core angle of the argument. Those changes produce genuinely new ads that reach fresh pockets of your audience.

How many new ads should a small business make each month?

Enough to release fresh creative every week and always have a tested replacement ready. For many small businesses that means a handful of genuinely different ads each week, drawn from a batch shoot every six to eight weeks. The exact number depends on budget and audience size, but a steady weekly rhythm matters more than a big quarterly drop.

AI referral traffic to United States retail sites grew 393% year over year in the first quarter of 2026, according to Adobe Analytics data reported by TechCrunch. March alone was up 269%. It is one of the fastest growing commerce traffic sources on the internet.

Here is the frustrating part. Most of that traffic lands on a page designed for a completely different kind of visitor, gets confused, and leaves.

Why do visitors from ChatGPT and other assistants bounce? Because they arrive looking to confirm one specific thing the assistant told them, and most pages open with something general instead. Put that fact in plain text at the top of the page and these visitors become some of the easiest sales you'll make.

Which is an odd thing to be annoyed about, because these are the best visitors you will get all year.

Who these people actually are

Someone arriving from a Google ad is early. They clicked a promise. They are still deciding whether the problem is worth solving.

Someone arriving from an AI assistant has already had a conversation. They described their situation, got options, asked follow ups, narrowed the field and then chose to come to you specifically.

The numbers reflect it. In Adobe's retail data, AI referred visitors converted 42% better than other visitors in March 2026 and spent 48% longer on site. We cover that gap in more detail in why traffic from AI converts better and almost nobody is tracking it.

They arrive knowing what you do, roughly what you charge and why they were sent. They are not browsing. They are verifying.

An AI referral is not a click. It is a warm introduction from a source your prospect already trusts. Most websites answer it with a hero banner and a slideshow.

Why they bounce anyway

The mismatch is simple. The assistant told them something specific. Your page opens with something general.

Say the assistant recommended you because you do same week commercial callouts across the Central Coast. Your visitor arrives looking for one thing, confirmation of that claim, and lands on a homepage that says "Excellence in service since 2011" above a photo of a handshake.

They now have to hunt for the fact that brought them. Plenty will not bother, because the assistant offered them two other options in the same breath and one of those might confirm faster.

The failure is not traffic quality. It is a verification page that never got built.

There's a second failure hiding underneath. Sometimes the assistant's description of you is out of date or slightly wrong, because the information it found about you was thin, old or inconsistent. The visitor arrives expecting a service, price or area you don't offer, and leaves disappointed. Fixing your page helps, but so does fixing what the wider web says about you. Our guide to what AI is telling your customers about you walks through how to check.

Building the page that closes them

Three changes handle almost all of it.

Lead with the specific claim, in plain text, above the fold. Not a headline about your values. The actual fact. Coverage area, turnaround time, price range, guarantee, credential. Whatever an assistant would use to justify recommending you, put it in the first thing a human reads.

Answer the follow up questions on the page. The prospect has already asked the assistant three or four clarifying questions. They arrive with a fifth. Put your genuine frequently asked questions on the page as real answers, not an accordion of one liners written for search engines in 2018.

Make the next step tiny and obvious. These visitors are close. They do not need nurturing. They need a phone number, a form with three fields and a clear statement of what happens after they submit it.

Plain text matters more than it sounds. Facts buried inside images, sliders or videos can't be read by the assistants that decide whether to recommend you, and they're slower for humans to find. The same sentence that confirms the claim for a visitor is the sentence a model can quote next time. One edit, two audiences.

How to actually see this traffic

This is where most businesses give up, because by default the report looks empty.

A large share of AI referrals arrive with no referrer at all and land in direct traffic. Some assistants pass a referrer, some strip it, and some route through link wrappers. So the fastest growing channel in your business hides inside the least informative bucket in your analytics.

Three fixes, in order of effort.

• Filter your referrals for assistant hostnames. Look for chatgpt.com, perplexity.ai, gemini.google.com, copilot.microsoft.com and openai.com in your referral report. Whatever shows up there is the visible tip of a much bigger number.
• Watch direct traffic behaviour, not volume. Segment direct visitors who land on a deep page rather than the homepage. Nobody types a service page URL from memory. Much of that segment is assistant referrals and dark social.
• Ask. Add one optional field to your enquiry form. "How did you hear about us." The number of people typing some version of "ChatGPT said" into that box will tell you more than any attribution tool you can buy.

Once you've found the segment, give it its own name in your reporting. A channel group or saved segment called "AI assistants" means you can watch it grow month by month instead of rediscovering it every time someone asks.

The compounding part

Here is what makes this worth acting on quickly rather than eventually.

Assistants are conservative recommenders. Once a model has a confident, well supported answer about who to hire in your category, it repeats that answer. It does not go looking for challengers every week.

So the businesses that get named now, and then confirm the recommendation with a page that closes cleanly, get named again. And the loop tightens, because satisfied buyers write reviews, which become more source material, which reinforces the recommendation.

The businesses that are ambiguous today do not get a second look next quarter. They get skipped by default.

What these visitors do differently on site

Watch a few session recordings from this segment and the pattern is unmistakable.

They scan rather than read. They ignore your navigation almost entirely. They hunt for one or two specific facts, usually price, coverage or timeframe, and they either find them in the first fifteen seconds or leave.

They also skip your carefully sequenced journey. The story you built across four scroll sections is irrelevant to someone who already had the story explained to them by a machine.

Design for scanning. Facts near the top, in plain text, in the order an assistant would have mentioned them. Save the persuasion for people who arrived without a briefing.

A worked example: the physio clinic's hidden channel

Here's a hypothetical. Say you run a physiotherapy clinic in Geelong. Your analytics show direct traffic creeping up and a handful of visits from chatgpt.com each month. Enquiries haven't moved, so you've never looked closer.

You build a segment of direct visitors who landed on deep pages and find most of them arriving on your "sports injury rehab" page. The session recordings show a pattern: they scroll straight past the hero image, look for something, and leave within seconds.

You ask a few assistants who they'd recommend for sports injury rehab in Geelong. You get named, alongside two competitors, with the reason given: Saturday appointments and direct health fund claiming. Your rehab page mentions neither in the first screen. Saturday hours live in the footer and health fund claiming is on a separate FAQ page.

So you rewrite the top of the page: "Sports injury rehab in Geelong, with Saturday appointments and on-the-spot health fund claiming. Book online in under a minute." Below that, you answer the questions people actually ask at reception, and the booking button sits right there.

Nothing about your marketing spend changed. You simply stopped making pre-sold visitors hunt for the reason they came. The details here are made up, but that's the shape of almost every fix in this channel.

Mistakes that waste AI referrals

• Sending every visitor through the homepage journey. Deep pages are where these visitors land, so deep pages need to stand on their own.
• Hiding prices entirely. If an assistant mentioned a price range, a page with no pricing signal at all looks evasive. A starting price or a range is often enough.
• Inconsistent facts across the web. Different hours or service areas on your site, Google Business Profile and directories confuse assistants and the people they send.
• Long forms. A ten-field form is a speed bump for someone who has already decided.
• Judging the channel on raw volume. This traffic is small and valuable. Measure it on enquiries and sales, not sessions. For the broader picture of pages that sell around the clock, see why your website should be your best salesperson.

What to do this week

Open your analytics. Find every session from an assistant hostname, and every direct session that landed on a deep page. Look at what those visitors did.

Then take your top three landing pages and rewrite the first fifty words of each so they confirm, in plain language, the specific fact that would have caused an assistant to send someone there.

That is an afternoon of work against one of the fastest growing traffic sources online, arriving pre sold, converting better, and currently bouncing off a slideshow.

There are not many fixes left in marketing with that ratio. Take this one. And if your site needs more than an afternoon of rewrites to get there, our web development team builds pages designed to confirm, answer and convert in the first screen.

Frequently asked questions

How do I track traffic from ChatGPT in Google Analytics?

Filter your referral report for assistant hostnames such as chatgpt.com, perplexity.ai, gemini.google.com and copilot.microsoft.com, and group them into their own channel. Because many AI referrals arrive with no referrer, also segment direct visitors who land on deep pages, and add a "How did you hear about us?" field to your enquiry form.

Why is my AI referral traffic bouncing?

Usually because visitors can't quickly confirm what the assistant told them. They arrive expecting a specific fact, like your service area, turnaround or price range, and land on a general hero banner. Put that fact in plain text at the top of the landing page, answer common follow-up questions below it and make the next step obvious.

Is AI referral traffic worth optimising for if the volume is small?

Yes. The volume is small compared with search, but the visitors arrive further along in their decision, having already compared options with an assistant. That makes each visit unusually valuable. The fixes are cheap too, mostly rewriting the top of key pages, so even a modest number of extra enquiries pays for the effort quickly.

How do I get ChatGPT to recommend my business?

Make clear, consistent, specific information about your business easy to find. State your services, area, pricing signals and credentials in plain text on your site, keep details matching across your Google Business Profile and directories, collect genuine reviews, and earn mentions on independent sites. Assistants recommend businesses they can describe confidently.

It cost them about four dollars and eleven minutes.

That is the world now. Copy is free. Design is free. Ad angles are free. Anything that can be described can be reproduced, and reproduced well enough that your prospect cannot tell the difference from a phone screen.

Which means every advantage built on execution has a shelf life measured in weeks.

There is exactly one asset left that cannot be cloned by a machine, and most businesses have never deliberately built it.

That asset is brand. Not the logo, the reputation: what people remember, say and search for when you are not in the room. When AI can write anyone's copy, brand is the only thing a buyer can't get from a cheaper clone, and it is also what AI assistants lean on when they decide who to recommend. Here is how to build it on purpose.

What a moat actually is

A moat is not a nice logo. It is not a colour palette or a font pairing or a tone of voice document nobody opens.

A moat is the reason a customer pays you more, waits longer, or chooses you without comparing. It is the gap between what you sell and what you are worth.

Historically businesses built that gap out of four things. Better product. Lower cost. Distribution. Or brand.

The first three are all under pressure. Product features get copied in a quarter. Cost advantages get competed away by someone with cheaper capital. Distribution keeps getting rearranged by platforms that owe you nothing and change the rules every eight months.

Brand is the one that compounds instead of eroding. And it is the one AI cannot manufacture, because brand is not what you say. It is what people remember and repeat when you are not in the room.

Anyone can generate your words. Nobody can generate your reputation. That is the entire game from here.

Why this got urgent in 2026 specifically

Two things collided.

The first is obvious. The cost of producing convincing marketing dropped to near zero, so the volume of convincing marketing went vertical. Your customer now sees more polished, more persuasive, more professionally worded messages per day than at any point in history, and trusts almost none of them.

The second is less obvious and more important. Discovery is moving into AI assistants, and assistants lean heavily on what the web says about businesses, not just what businesses say about themselves.

Think about what that does. When a buyer asks an assistant who they should hire, the answer is assembled from reviews, mentions, forum threads, articles and the general shape of your reputation across the internet. Your beautifully written homepage is one weak vote among many.

The businesses with a real brand get named. The businesses with good marketing and no reputation get summarised into a category and skipped.

The three ingredients, in order

One. A position, not a description. Most businesses describe what they do. A position says who you are for, who you are not for, and what you refuse to compromise on.

"We do web design for small business" is a description. "We build sites for trades who need the phone to ring, and we will not take a project without a photography budget" is a position. The second one loses you clients. That is what makes it work.

Two. Consistency long past the point of boredom. Brand is built by repetition, and repetition feels like failure to the person doing it. You will be sick of your own message long before the market has noticed it.

Most businesses that have built genuine recognition went through that valley. The ones that rebranded to relieve their own boredom reset the clock to zero and called it a fresh start.

Three. Proof that lives outside your website. Named results. Real customers on camera. Case studies with numbers in them. Reviews with detail. Mentions in places you do not control.

This is the ingredient that does double duty. It builds trust with humans and it is exactly the material AI assistants read when deciding who to recommend.

If you have never looked at what that material currently says, start there. Ask a few assistants who they would recommend in your category and what they know about you. We walk through that exercise in your customers are asking AI about you. Reviews carry more of the weight than most owners expect, which is why reviews are the new backlinks.

The test that tells you if you have one

Take your marketing. Remove your logo, your name and your colours. Show it to someone in your industry.

Can they tell it is you?

For most businesses the honest answer is no, because the copy could belong to any of forty competitors and the imagery came from the same stock library everyone else uses.

Here is a harder version. Ask three recent customers to describe what you do, in their own words, without prompting. If you get three different answers, you do not have a brand. You have a business that people like, which is not the same thing and does not compound.

What building it actually looks like

It is less glamorous than a rebrand and considerably cheaper.

• Pick a fight. Name the thing your industry does badly and refuse to do it. This gives customers a reason to choose you that has nothing to do with price.
• Put a real human at the front. A recognisable face beats a stock model every single time, and it is the one visual asset that cannot be generated. Same person, same energy, across everything.
• Say the same thing for two years. One core promise. Repeated in the ads, on the site, in the emails, on the invoices, in the way the phone gets answered.
• Publish proof relentlessly. Every job with a number attached is an asset. Plenty of businesses complete dozens of them a year and document none.
• Go and get mentioned. Suppliers, industry publications, local media, podcasts, roundups. Third party mentions are the currency of both human trust and AI recommendation.

A worked example: two electricians, one suburb

Here is a hypothetical. Two electricians work the same Sydney suburbs. Similar prices, similar skills, similar websites that both say "reliable, licensed, affordable".

The first keeps doing what he has always done. Good work, happy customers, no system for capturing any of it. When a homeowner asks an assistant for a recommended local electrician, he rarely comes up, because there is almost nothing about him online beyond a handful of short reviews. He wins jobs on price and on being available this week.

The second makes three deliberate decisions. She picks a position: switchboard upgrades and EV charger installs for older homes, and she will not do cheap patch jobs on unsafe wiring. She puts her own face on the van, the site and every short video explaining a common problem in plain English. And after every job she asks for a detailed review and photographs the before and after, with the owner's permission.

Two years later her enquiries arrive already sold. People say "I saw your video about old switchboards" on the first call. Price barely comes up. Local groups tag her when someone asks for help. Assistants have a clear, consistent picture of who she is and what she does. Same trade, same market, same tools. One has a business people like. The other has a brand people repeat.

How to measure a brand you cannot screenshot

Brand feels fuzzy until you pick the right signals. None of these will spike overnight. All of them trend, and the trend is the point.

• Branded search. Are more people searching for your name specifically? Search Console shows this for free.
• Direct and referral enquiries. Track how every lead found you. A rising share of "a friend told me" and "I already knew about you" is brand compounding.
• Price objections. Count how often price is the first question on a sales call. As brand grows, it should come up later and less.
• Close rate. Pre-sold prospects close faster and more often. If your close rate climbs while your ad spend stays flat, brand is doing the work.
• Unprompted descriptions. Repeat the three-customer test every six months. You want the answers converging on the same words.

The mistakes that reset the clock

Confusing a rebrand with brand building. A new logo is a cost. Brand is the reputation that builds up behind a consistent name over years. Changing the look to feel fresh throws away recognition you already paid for. Get the positioning right before any visual work, as we argue in positioning comes before the logo.

Chasing every trend. Jumping on each new format and voice makes you look like everyone else doing the same. Pick the channels where your customers are and show up there consistently.

Hiding the humans. Faceless businesses are easy to swap. The founder, the team and real customers are the parts of your brand a competitor cannot copy.

Letting proof evaporate. Every completed job without a review, a photo or a documented result is brand equity you earned and then threw away.

The uncomfortable maths

Brand is slow. It does not produce a spike you can screenshot at the end of the month. It produces a gradual, almost invisible shift where more of your enquiries arrive already sold, price objections get quieter, and referrals start turning up from people you have never met.

That shift is worth more than any campaign, and it is the only thing on your balance sheet that gets more valuable while everything else gets easier to copy.

Your competitors will keep buying tactics. Ad platforms will keep changing. AI will keep making execution cheaper for everybody, including the people trying to take your customers.

Build the one thing that does not care.

That is the work we do in consulting at LIQID NOISE: finding the position you can own, the proof you already have, and the message you can repeat for years without flinching. It is slower than a campaign. It is also the only marketing that gets more valuable every year you keep doing it.

Frequently asked questions

What is a brand moat?

A brand moat is the reason customers choose you, pay more or wait longer without shopping around. It is built from reputation, recognition and trust rather than features or price, which is what makes it hard to copy. A competitor can clone your website overnight. They cannot clone years of reviews, referrals and people repeating your name.

Why does brand matter more now that AI can write marketing copy?

Because AI makes execution cheap for everyone. Polished copy, design and ad angles can be reproduced in minutes, so they no longer set you apart. What remains scarce is reputation: what customers say, what reviews describe and what third parties mention. AI assistants also draw on that material when recommending businesses, so brand now shapes discovery too.

How long does it take to build a brand?

Longer than most businesses expect. Recognition builds through repetition, and you will usually be tired of your own message well before the market has noticed it. Expect to hold one core promise, one face and one consistent look for years, not months. The early signs are more pre-sold enquiries and fewer price objections.

Can a small business build a strong brand without a big budget?

Yes. Brand building for a small business is mostly consistency and proof, not spend. Choose a clear position, put a real person at the front, collect detailed reviews after every job, document results and pursue mentions in local media, industry groups and podcasts. These cost time more than money, and they compound.

Clicks down 30%. Impressions flat or up. Rankings unchanged.

Nothing broke. Nothing was penalised. Google simply started answering the question on the results page. Pew Research Center found that Google users who saw an AI summary clicked a traditional result in 8% of visits, against 15% when no summary appeared.

So why is your Google traffic down when your rankings aren't? Because AI answers now satisfy a big chunk of searches before anyone clicks. You get paid anyway by owning the questions a machine can't settle in a sentence, becoming the source those answers quote, and building channels Google doesn't sit between.

Seer Interactive found organic click-through rates on informational queries with AI Overviews fell 61% between June 2024 and September 2025. How hard any one business gets hit depends almost entirely on one thing, which we will get to.

First, the part nobody wants to say out loud. That traffic is not coming back. Not with better titles, not with more posts, not with a technical audit. The interface changed permanently.

The revenue, though, is absolutely recoverable. It just stops coming from where you are used to looking.

What actually changed on the results page

For twenty-odd years, search worked like a directory. You typed a question, Google showed you ten links, and you clicked one to get the answer. Every click was a visit, and every visit was a chance to win a customer.

AI Overviews and AI Mode flipped that. Now Google reads the pages for the searcher, writes a summary, and puts it at the top. Links still appear, tucked inside or beside the answer, but the searcher often has what they need before they ever reach them. Your page can be doing the work of informing the answer and still get no visit for it.

That's why the Search Console pattern looks so strange. Impressions hold up or grow, because your page is still being shown. Rankings look fine, because you're still in position. Clicks fall, because the job the click used to do is being done on Google's page instead of yours.

Understanding that mechanism matters, because it tells you what not to do. If the problem isn't your rankings, chasing rankings won't fix it.

Why some businesses lost a little and others lost most of it

The businesses that got flattened were the ones whose traffic was built on questions a machine can answer in a sentence.

"What temperature to serve red wine." "How many square metres in an acre." "What is a good conversion rate." That traffic was never buying traffic. It was curiosity traffic that occasionally converted by accident, and Google absorbed it entirely.

The businesses barely scratched were the ones ranking for questions where the answer is a decision, not a fact. "Should I repoint or replace." "Which of these two suppliers handles regional callouts." "What goes wrong when you do this cheap."

If a sentence can answer the query, Google will answer it. If a judgement is required, the click survives. Audit your top pages against that one line and you will know exactly what is coming.

Run that test on your own top twenty pages this week. Split them into fact pages and decision pages. The fact pages are already gone or going. The decision pages are the asset you should be defending and expanding.

Stop counting sessions, start counting the three that matter

Traffic was always a proxy metric. It stopped being a useful one the moment discovery moved off your site.

Three numbers replace it, and all three are available to you today without any new software.

• Branded search volume. The number of people typing your business name into Google. When an assistant recommends you, this is where it shows up. It is the closest thing to an honest scoreboard that exists right now.
• Direct traffic and its quality. Not the number. The behaviour. Direct visitors who land, read one page and enquire were sent by something. Usually a conversation you cannot see.
• Enquiry quality. The most underrated metric in marketing. Fewer, better briefed leads is what the new search environment produces, and most businesses record it as a bad month.

Plenty of marketers now describe the same pattern: search volume down while the intent behind each search is higher, because buyers are further along the journey by the time they type anything. Fewer sessions, warmer people. If you judge that shift on a traffic chart, you will cut the channel that is actually working. We dig into that shift in why fewer searchers are more ready to buy.

To track these properly, set up three simple habits. Check branded queries in Search Console every month and chart them over time. Build a segment in your analytics for direct visitors who land on a deep page rather than your homepage. And add a "How did you hear about us?" field to your enquiry form, then read the answers. Together, those three tell you more about how discovery is working than any traffic report.

The three plays that still bring in money

Play one. Own the decision content. Comparisons, buying criteria, cost breakdowns, what goes wrong, how to choose. These pages survive because the answer depends on the reader's situation, and an AI summary that tries to answer them still has to send the reader somewhere for the detail.

Write them properly. Real numbers, real trade offs, an honest section on when you are the wrong choice. That last part is what makes both humans and models trust the page.

Play two. Get quoted instead of clicked. If Google is going to summarise the answer, be the source it summarises. The research on this points one way. The GEO study presented at KDD 2024 found that adding citations, statistics and quotations to a page lifted its visibility in AI generated answers by roughly 30% to 40% on the researchers' own test engine. Treat that as a direction, not a guarantee.

Practically, that means every important page leads with a direct answer in the first two lines, contains numbers rather than adjectives, and quotes named humans. A page that reads like a well sourced briefing gets cited. A page that reads like a brochure gets ignored. Our GEO playbook for getting cited by ChatGPT, Gemini and Perplexity covers the page-level detail.

One warning here. Ranking well on Google no longer guarantees you'll be one of the sources an AI answer draws from. The two lists overlap less than most people assume, which we explain in why ranking number one no longer gets you into the answer. You have to earn the citation on its own terms.

Play three. Build the channels Google cannot mediate. Email list. Customer database. Reviews. Referral relationships. A physical presence in your community. None of this is exciting and all of it becomes more valuable every single quarter that discovery gets more automated.

A worked example: the builder with two kinds of pages

Here's a hypothetical to make the audit concrete. Say you run a renovation business in Wollongong. Your site has built up a library of blog posts over the years. Your five biggest traffic pages are things like "how long does grout take to dry" and "what is a load-bearing wall." Your smaller pages include "bathroom renovation cost in Wollongong" and "renovate or knock down and rebuild: how to decide."

The traffic drop lands almost entirely on the first group. Those were fact pages. Google now answers them directly, and the people asking were mostly DIYers who were never going to hire you.

The second group barely moves, because those questions need judgement about the reader's situation. So you redirect your effort. You rewrite the cost page to open with a direct answer and a realistic range based on your own jobs, add the factors that push a project up or down, and include an honest section on when a cheaper trade is fine. You do the same for the renovate-or-rebuild page, with a named quote from you on the mistakes you see most.

Total traffic might keep drifting down. But enquiries from those decision pages arrive better briefed, and the people who call already understand the ballpark. That's the trade the new search environment offers, and it's a good one if you take it deliberately.

What to actually do this month

Pick your five highest intent pages. The ones tied to money, not traffic.

Rewrite the opening of each so the first sentence directly answers the question in the title. Add three specific numbers you can defend. Add one named quote, from you, a customer or a source you cite properly. Add a section on how to choose between options, including when a competitor is the better call.

Then go and get mentioned somewhere you do not control. A supplier's blog. An industry roundup. A local publication. A genuinely useful answer in a forum where your customers actually ask questions. Assistants weight independent mentions heavily and most of your competitors have none.

That is a fortnight of work and it does more for your visibility than another twelve blog posts aimed at a search results page that no longer sends clicks.

Mistakes that make the drop worse

• Publishing more fact content to make up the numbers. You're feeding the exact type of page Google now answers itself.
• Deleting pages because traffic fell. A page with fewer clicks may still be informing AI answers and supporting your authority. Judge pages on enquiries and assisted conversions, not sessions alone.
• Blaming a penalty that doesn't exist. If rankings and impressions hold while clicks fall, a technical overhaul won't bring the clicks back.
• Burying the answer. Long intros before the point make your page harder to quote and more annoying to read.
• Ignoring your reputation off-site. Reviews and mentions on other sites feed the answers too. Our piece on how AI reads your reviews before it recommends you covers why.

The bigger reframe

The blue link era rewarded whoever could produce the most pages. The answer era rewards whoever is most worth quoting.

That is a harder game, and a much better one for any business with real expertise and actual results. The content mill cannot win it. You can.

If you want help auditing your pages, rebuilding the ones that matter and tracking what actually brings in revenue now, that's exactly what our digital marketing team does for search in the AI era.

Stop mourning the sessions. Go and become the source.

Frequently asked questions

Why is my organic traffic down if my rankings haven't changed?

Most likely because Google's AI Overviews and AI Mode now answer the query directly on the results page. Your page is still shown, so impressions and rankings hold, but fewer people need to click through. Check Search Console: flat or rising impressions with falling clicks and a stable average position is the classic pattern, not a penalty.

Can I stop Google's AI Overviews from using my content?

Blocking your content from AI features usually costs more visibility than it saves, because the same systems increasingly decide what searchers see. For most businesses, the smarter move is to become the source those answers cite, by leading pages with direct answers, adding specific numbers and named quotes, and earning mentions on other trusted sites.

What kind of content still gets clicks from Google in 2026?

Content that helps someone make a decision rather than look up a fact. Comparisons, cost breakdowns, buying criteria, what goes wrong and how to choose between options all depend on the reader's situation, so a summary can't fully replace them. Honest, specific decision pages still earn clicks and tend to send better qualified enquiries.

How do I measure SEO success when traffic is falling?

Track branded search volume, the behaviour of direct visitors who land on deep pages, and the number and quality of enquiries. Add a "How did you hear about us?" field to your forms. Rising branded searches and better briefed leads can mean search is working harder for you, even while total sessions go down.

Plenty of advertisers took that literally. They built accounts with 395 live ads. Then they watched performance flatten and could not work out why more was suddenly doing less.

Here is what happened. Meta's Andromeda ranking system got very good at recognising when two ads are the same ad wearing a different hat.

Andromeda clusters near identical creative into a single entity. Three hundred and ninety five ads that share a hook, a format and a visual language do not get treated as 395 chances. They get treated as roughly one. You paid for 395 exports. You bought one shot. That is why fifteen genuinely different ads can now beat three hundred lookalikes: the system rewards distinct ideas, not file counts.

The number that gets misread

You have probably seen the benchmark claiming the best performing advertisers run far more live ads than the weakest. Every agency deck in the country screenshotted that stat and turned it into "make more ads."

They read the wrong half of the finding. The top performers are not winning because of the count. They are winning because their ads contain genuinely different ideas, while weaker accounts often contain a handful of ideas duplicated over and over.

Andromeda does not reward how many files you uploaded. It rewards how many different ways you gave it to reach a human.

The correction is not to stop producing. Common Thread Collective reports that brands testing 20 or more new ads per month still see 65% higher return on ad spend than brands testing fewer than 10. Volume matters. Volume of sameness does not.

What counts as a different ad

This is where most accounts fall over, because "different" gets defined as "different colour grade." Under Andromeda, difference has to be structural. Four dimensions decide whether the system sees two separate concepts or one.

Hook. The first three seconds and the first line of copy. A price hook, a problem hook, a myth busting hook and a founder story hook are four ads. Four crops of the same product shot are one.

Format. Static, carousel, talking head, screen recording, UGC selfie video, motion graphic, before and after. Each is a genuinely different rendering pipeline in the model's eyes.

Presenter and setting. A customer in their kitchen and a studio product shot are not the same ad even if the script is identical. Faces, rooms and lighting are exactly the signals the clustering is reading.

Angle. The actual argument. Speed versus price versus risk reversal versus social proof versus fear of getting it wrong. This is the one that separates real creative strategy from asset production.

Just as useful is knowing what does not count. A new colour grade. A different font. A swapped music bed. A tighter crop. New wording on the call to action button. A different stock background behind the same product. These are finishing decisions. They can still matter at the margins, so test them inside a concept if you want to, but never count them as new concepts and never let them fill a production schedule that should be spent on new hooks and angles. If a variant would take a designer ten minutes to make, it is almost certainly not a different ad in the system's eyes.

The structure that goes with it

Creative diversity only pays if the account is built to let the system move budget freely. Three things consistently separate the accounts that got the Andromeda uplift from the ones that got a shrug.

• Fifteen to twenty genuinely distinct live ads. Not eighty variants. Fifteen real concepts, refreshed constantly, is the shape that works.
• Consolidated campaign structure. Most accounts should have the bulk of spend running through Advantage+ Shopping rather than a museum of legacy ad sets that each hold a fraction of the data.
• Clean signal. Pixel and Conversions API running together, with Event Match Quality above 7. Andromeda is a prediction engine. Feed it noisy conversion data and it predicts badly, no matter how good the creative is.

Much of the lift from moving off legacy setups to an Andromeda aligned structure is not a creative gain. That is the same creative, given a system that can actually use it.

The clean signal point is the one most often skipped, because it is invisible in the ads manager dashboard until you go looking for it. If you have never checked your Event Match Quality, start there. Our guide to the boring number quietly deciding your ad results walks through what it measures and how to raise it.

The fatigue trap underneath all of this

There is a second change that makes the diversity problem urgent rather than academic. ScaledOn's analysis puts effective ad lifespan at two to four weeks now, down from six to eight before Andromeda.

Do the arithmetic on that. If your winners burn out in 14 to 21 days, and you produce creative in quarterly bursts, you spend most of every quarter running ads the audience has already stopped seeing. Then you conclude that Meta stopped working.

Meta did not stop working. Your production cadence stopped matching the platform's metabolism.

The two changes compound each other. Fatigue means you need fresh creative more often. Clustering means that fresh creative only counts if it is genuinely new. An account that refreshes every fortnight with recoloured versions of last month's winner gets the worst of both: the audience is tired of the idea, and the system does not see anything new to test. We go deeper on the fatigue side in your winning ad now dies in 14 days.

How to build the pipeline without a studio

You do not need a production company. You need a repeatable day.

One shoot day, structured properly, produces a quarter of distinct creative. Book a half day with a real customer or two, a phone, a clip mic and a shot list built around angles rather than assets. Capture four hooks, three formats and two settings and you walk out with 24 genuinely different combinations before a single edit.

Then edit against the four dimensions above, not against a template. If two exports differ only by font, delete one. It was never going to get its own audition anyway.

If you want the full playbook for turning one day on set into months of material, we break it down in one video shoot, 90 days of content.

Ship in weekly batches rather than monthly dumps. Weekly gives the system a rolling supply and gives you a clean read on which angle is actually carrying the account.

How to tell if your account has a sameness problem

You do not need special tools to diagnose this. You need an honest hour with your live ads.

1. Export every live ad. Put the first frame and the first line of copy for each one side by side.
2. Tag each ad on the four dimensions. Hook, format, presenter and setting, angle. Be strict. A new headline on the same argument is the same angle.
3. Count the unique combinations. Not the ads. The genuinely different concepts. This is the number Andromeda is effectively working with.
4. Look at where spend is going. If one or two concepts are soaking up almost all the budget while dozens of near duplicates get scraps, the system has already clustered them for you.
5. Check the dates. How old is your top spending concept? If it has been running well past the two to four week fatigue window, you are living on borrowed time.

Most accounts that run this audit find the same thing: a long list of ads and a short list of ideas. That gap is the work.

A worked example: the skincare brand with 120 ads

Say you run a hypothetical Melbourne skincare brand spending $15,000 a month on Meta. You have 120 live ads. Tag them and you find the truth: nearly all of them are the same product shot on a marble background with a "clean ingredients" message, in different crops, fonts and colours. Three concepts, dressed up as 120.

The rebuild starts with the list of reasons people buy. Sensitive skin that reacts to everything. A routine that takes too long. Scepticism about "natural" claims. Wanting to look less tired. Distrust of influencer hype. Each becomes an angle. Then you book a half day with two real customers, a phone and a clip mic. You capture a founder story hook, a myth busting hook about ingredients, a problem hook about reactions and a routine-in-sixty-seconds hook, across talking head, selfie UGC and screen-recorded product demo formats, in a bathroom and a kitchen.

You cut 120 ads down to about fifteen genuinely distinct concepts, consolidate campaigns, fix the Conversions API setup, and ship three or four new concepts every week. Nothing here guarantees a result. But the account now gives the system fifteen different ways to find a buyer instead of three, and gives you a clean read on which angles are actually selling.

The uncomfortable part

Andromeda made creative strategy the constraint, and creative strategy is the one part of paid media you cannot buy off a shelf or automate away.

Bidding is automated. Targeting is automated. Placement is automated. What is left is knowing your customer well enough to have fifteen genuinely different true things to say to them.

Most businesses do not have fifteen. They have three, and eleven ways of restating the second one.

That is the real work. Sit down and write out every distinct reason a person buys from you, every objection that stops them, and every proof you can put on camera. When that list gets past twenty, your Meta account fixes itself.

That is also where outside help earns its fee. Not in pushing buttons in ads manager, which the platform increasingly does for you, but in extracting those twenty true things from your business and turning them into distinct creative every week. Our advertising team builds that pipeline: the angle list, the shoot days, the edit against the four dimensions, and the account structure that lets Andromeda actually use it.

Frequently asked questions

What is Meta Andromeda?

Andromeda is Meta's ad ranking system, the part of the machine that decides which ads are worth showing to which people. Its practical impact for advertisers is that it recognises near identical creative and groups it together. Ads that share a hook, format and visual style are treated as roughly one concept, so real creative diversity matters more than ad count.

How many ads should I run on Meta after Andromeda?

Aim for fifteen to twenty genuinely distinct live ads rather than dozens of variants. Distinct means a different hook, format, presenter or setting, or angle, not a new colour or font. Keep refreshing with new concepts, since Common Thread Collective reports brands testing 20 or more new ads a month see 65% higher return on ad spend than those testing fewer than 10.

Does changing the colour or headline make a new ad for Meta?

Usually not. Cosmetic changes such as colour grade, font, crop or minor headline tweaks leave the core concept intact, so the system is likely to treat the variants as the same ad. To register as different, change something structural: the opening hook, the format, the person and place on screen, or the actual argument you are making to the buyer.

Should I still use Advantage+ campaigns?

For most accounts, consolidating spend is the right move, with the bulk of it running through Advantage+ Shopping rather than many small legacy ad sets. Consolidation gives the system more data per campaign and more freedom to move budget toward what works. It only pays off when paired with diverse creative and clean conversion signal from the Pixel and Conversions API.

That was shopping in 2019.

Here is shopping in 2026. Your customer types one sentence into an assistant. "I need a commercial coffee machine for a 40 seat cafe in Newcastle, under twelve grand, serviced locally." Four seconds later they have three options, a comparison of warranty terms, and a recommendation.

They did not visit your website. They did not read your about page. They did not see your Google reviews. A machine did all of that on their behalf and then handed them a shortlist. Whether you make it comes down to three things: consistent facts, specifics a machine can quote, and what other people say about you.

In a Riskified survey of 5,400 shoppers across nine markets including Australia, run in late 2025, 73% said they already use AI somewhere in their shopping journey. That is not a future trend. That is right now, in your category, today.

So the only question that matters is brutally simple. When the machine builds that shortlist, is your business on it?

The shortlist is the new shopfront

For twenty five years the goal was to be found. Rank on page one, run the ad, get the click, then convince them on your website.

Agentic commerce breaks that chain in the middle. The assistant does the finding, the filtering and most of the convincing. Your website is no longer where the sale is decided. It is where the sale is confirmed.

Look at the shape of it. ChatGPT already surfaces products from merchants in its answers, and since OpenAI reworked its shopping features in 2026, buyers complete the purchase on the merchant's own site. Perplexity, Gemini and Amazon's Rufus are running their own versions. Discovery happens inside the assistant. Checkout happens on the merchant's site.

Which means you are not competing for a click any more. You are competing to be one of three names in a paragraph.

Being on page one used to mean being one of ten. Being in the answer means being one of three. The middle of the market just disappeared.

Why the machine picks one business over another

Assistants are not magic and they are not random. They assemble an answer out of whatever the open web says about you, cross referenced against whatever structured data they can read. Three things decide whether that is flattering.

One. Whether your facts are consistent. If your website says you service the Central Coast, your Google profile says Sydney, and a directory from 2021 says you closed, the model has three contradictory facts and no way to resolve them. Contradiction is the single fastest way to get dropped from a recommendation, because a model that is unsure will simply name a competitor it is sure about.

Two. Whether your specifics are extractable. "We pride ourselves on quality workmanship and customer service" is unquotable. It contains no fact a machine can lift. "We service commercial espresso machines across the Central Coast and Newcastle, with a 24 hour callout guarantee and a two year parts warranty" is a sentence an assistant can repeat word for word. One of those businesses gets named. The other does not.

Three. What third parties say. Assistants weight independent sources heavily. Reviews, forum threads, roundups, industry directories and comparison articles are where they go looking for confirmation. If nobody outside your own website has ever described what you do, you are asking a machine to trust your marketing copy alone. It will not. Reviews in particular do more work here than most owners realise, which we unpack in AI reads your reviews before it recommends you.

How to make your business easy to recommend

Knowing why machines choose is useful. Knowing what to change on Monday is better. Here's the order that works for most businesses.

1. Write your one-line fact sheet. What you sell, who for, where, at what price range, with what guarantee. Plain sentences, no adjectives. This becomes the source of truth for everything else.
2. Make every listing match it. Website, Google Business Profile, social bios, marketplaces and industry directories. Same name, same service area, same description. Hunt down the old listings too, because stale pages are where contradictions hide.
3. Rewrite key pages for extraction. Put the specific facts in the first lines of your homepage, service pages and product pages. Every page should contain at least one sentence an assistant could quote on its own.
4. Add structured data. Business details, products or services, prices, availability, reviews and FAQs, marked up with schema so machines don't have to guess.
5. Publish comparison and buying-guide content. Assistants field questions like "which is better for a small cafe, X or Y?" A clear, honest comparison page on your site gives them something to draw from.
6. Earn outside mentions. Ask customers for reviews that mention the specific job or product. Get listed in the directories your buyers trust. Contribute useful articles to industry publications.
7. Check you're reachable. Make sure your site isn't blocking the crawlers and agents you want to read it. For product sellers, the technical side is covered in the protocols deciding which brands AI agents can buy from.

The four second audit you can run right now

Stop reading. Open ChatGPT, Gemini or Perplexity. Type the exact sentence one of your customers would type. Not your brand name. The problem they have.

"Best [what you do] in [your city]." "Who should I hire to [the outcome you sell]." "Compare [your category] providers in Australia."

Then read the answer like a customer would. Three things usually happen, and each one tells you what to fix.

• You are not mentioned. The model has no confident entity for your business. Fix the fundamentals first, consistent facts everywhere, real structured data, and third party mentions that are not paid ads.
• You are mentioned but described wrong. Old pricing, a service you dropped, a suburb you left. The model is quoting a stale internet. Update the sources, not just your homepage.
• A competitor is described better than you. Read their description closely. It will contain specifics yours does not. That gap is your entire content brief for the next quarter.

For a fuller version of this audit, including how to trace each error back to the page that caused it, see what AI is telling your customers about you.

A worked example: the coffee machine shortlist

Go back to that cafe owner in Newcastle. Say you're one of the businesses that could serve her. You sell and service commercial espresso machines from a workshop in Gosford. You've been doing it for years and your customers love you.

She types her sentence. The assistant names three suppliers. You're not one of them.

You go looking for why. Your homepage headline says "Passion for the perfect pour." Your service area is mentioned once, in the footer. Your Google profile lists Gosford but not Newcastle. An old directory entry still shows a number you disconnected. Your price ranges live in a PDF. You have plenty of happy customers but only a handful of reviews, and none mention Newcastle or callout times.

Every one of those is fixable in a few weeks. You rewrite the homepage opening: "We sell and service commercial espresso machines for cafes across the Central Coast and Newcastle, with a 24 hour callout guarantee." You update every listing to say the same. You move price ranges into plain text on the product pages and mark them up. You ask your last thirty cafe customers for a review that says where they are and how fast you turned up.

Nothing about the business changed. What changed is that a machine can now read it, confirm it and repeat it. That's the difference between being known locally and being named in the answer.

What this actually changes about your marketing

It changes what you publish. Vague brand copy has always been weak. Now it is invisible. Every page needs at least one line a machine can lift and repeat without needing the paragraph around it.

It changes what you measure. If a buyer spends nine minutes in an AI conversation and then types your business name into Google, your analytics records a branded search out of nowhere. That is not luck. That is the assistant doing your selling, and you need to start counting it.

Practically, that means a few new habits. Set up a separate channel in your analytics for visits referred by AI tools. Watch branded search volume month to month. Add "an AI assistant" as an option in your "how did you hear about us?" question. And keep a saved list of buyer prompts you re-run every month. Visitors who arrive from AI tend to be further along in their decision, a point we explore in why AI traffic converts better.

It changes what you invest in. A single strong comparison page, a well structured service page and thirty genuine reviews will do more for your AI visibility this year than another rebrand.

And it changes the cost of doing nothing. Every month you stay ambiguous, the assistants get more confident about the competitor who is not.

The mistakes that keep businesses off the shortlist

• Only checking your brand name. Buyers rarely ask about you by name. They ask about their problem. Audit the problem prompts.
• Hiding facts in PDFs and images. Price lists, warranty terms and service areas locked in a brochure are invisible to most machines.
• Fixing the homepage and stopping. The model reads the whole web, so old listings and profiles need fixing too.
• Chasing tricks. Hidden text, keyword stuffing and fake reviews don't build a confident entity. They build risk.
• Assuming it's only for ecommerce. Service businesses get shortlisted, compared and booked through assistants too.

The window is open, briefly

Here is the good news. Almost nobody in your category has done this work. Most businesses still have not typed their own category into an assistant to see what comes back.

The businesses that fix their facts, publish extractable specifics and go earn third party mentions right now will be the ones the models learned first. Models are conservative. Once they have a confident answer, they keep giving it.

Being early is not a nice advantage here. It is the whole advantage.

If you'd like help doing it properly, this is exactly what our digital marketing team works on: fact cleanup, extractable pages, structured data, reviews and the monthly prompt checks that show whether it's working.

Go type the question. See what the machine says about you. Then decide whether you are comfortable letting that answer stand.

Frequently asked questions

What is agentic commerce?

Agentic commerce is shopping where an AI assistant or agent does much of the research on the buyer's behalf. The buyer describes what they need, and the assistant finds options, compares them and recommends a shortlist. In the current model, discovery happens inside the assistant and checkout usually happens on the merchant's own site, so the sale is often decided before the buyer visits you.

How do AI shopping assistants choose which businesses to recommend?

They draw on what the web says about you and what structured data they can read. Businesses with consistent facts across every listing, specific quotable details on their pages and strong independent confirmation from reviews and third-party sources are easier for a model to recommend with confidence. Contradictory or vague information makes a model more likely to name a competitor it is surer about.

Do service businesses need to worry about AI agents?

Yes. Agents don't only buy products. They shortlist tradespeople, compare consultants and help book appointments. For a service business, the equivalent of product data is service data: what you do, where, for whom, at what price range and with what guarantee. Publish that clearly and consistently and you're doing the same work as an online store.

How do I track sales that come from ChatGPT or other AI tools?

Set up a separate channel in your analytics for referrals from AI tools, watch branded search volume over time and add AI assistants as an option in your "how did you hear about us?" question. Many buyers research in an assistant, then search your name directly, so rising branded searches can be a sign the assistants are recommending you.

Musicians are asking it. Designers are asking it. The Guardian ran a full essay on it. Your group chat has probably debated it twice this month.

So let me give you the short answer up front. No. AI is not going to kill art.

But something else is happening while everyone argues about the wrong question. AI is quietly splitting every creative field into two groups. People who can only generate. And people who can actually draw, design, and see. One of those groups is about to be worth nothing. The other is about to become the most expensive talent on the market.

Which group you land in is a choice you are making right now, today, with how you spend your practice hours. Let me show you why.

The panic is real, and it's not stupid

In 2024 the Guardian published a piece by musician and writer Rudi Zygadlo titled "Should artists be terrified of AI replacing them?" It opens with him typing his own creative dread into an image generator and getting back something that looks like a holiday ad. Then he says the line that every working creative has felt in their gut: "AI has killed my bucket list."

Why write the album, the screenplay, the novel, when a model can spit out a passable version before your coffee cools? Why spend ten years learning to paint when a prompt takes ten seconds?

If you've felt that vertigo, you're not weak and you're not a luddite. You're paying attention. Every previous tool, from the fountain pen to the word processor, was something an artist held. As Zygadlo points out, AI is different because it threatens to be both the tool and the artist.

So he did something smart. He took the question to the artists who live deepest inside the machine.

The verdict from the people who built art with AI first

Holly Herndon and Mat Dryhurst have been working with AI, and teaching people about it, for more than a decade. They trained an AI "baby" to sing on Herndon's 2019 album Proto. If anyone had a motive to announce the death of the human artist, it's the two people who got there early.

Zygadlo asked them point blank: is AI going to kill the human artist? Their answer was one word. No. And then they gave the reason, and it's the single most clarifying sentence in this entire debate:

"A media file that sounds like a choir is not a choir. Culture will persist and evolve in unexpected ways."

Read that again. The file is not the thing. The output is not the art. And their proof is something you've already paid money for: DJing. On a technical level, DJing is very easy to automate. Software can beat-match for you. And yet people pack clubs and drop hundreds on festival tickets to watch a person do it. Because nobody was ever buying the audio. They were buying the human, the room, the taste, the moment.

Writer Luke Fewtrell landed in the same place in his Medium essay "Will AI Kill Art", and he compressed it into seven words: "Art only dies when the artist dies." Humans have been making art since we were grinding pigment onto cave walls. Every technology panic since has followed the same script. The camera was supposed to kill painting. Painting is still here. Synthesisers were supposed to kill musicians. The Buggles wrote "Video Killed the Radio Star" and radio outlived the music video. Art survives every single time because making it is not a market activity. It's a human need. As Fewtrell puts it, the feeling of creating something born from your own mind is unlike anything else.

So the comfortable take is true. Art will survive. But I need you to hear the uncomfortable part, because "art will survive" is not the same as "your career will survive."

AI doesn't kill art. It kills the average.

Here's what generative models actually are. They're trained on the sum of everything that already exists, and they produce a statistical blend of it. That makes them, by definition, machines for manufacturing the middle. Competent, polished, instantly familiar, and utterly interchangeable.

Now apply first-grade economics. When anyone can generate a decent illustration in ten seconds for free, what is a decent illustration worth? Exactly what it costs to make. Nothing. The floor of the creative market, the "pretty good and fast" tier, is being priced at zero as you read this. I wrote about what that sameness does to brands in my breakdown of Meta's AI product photos, and audiences already have a name for the flood: AI slop. They scroll past it on instinct.

So AI is not coming for art. It's coming for average art. It's coming for the person whose entire value was "I can produce something decent-looking on demand." That skill is now a utility, like tap water.

Which raises the only question that matters: what does the market pay for when decent is free?

Why the pencil just became a power tool

It pays for the things a prompt cannot produce. And every one of them comes from old-fashioned, unglamorous, hand-earned craft.

Drawing teaches you to see. When you learn to draw, the drawing is almost a byproduct. The real product is your eye. You learn why a composition feels balanced, why light behaves the way it does, why a face reads as trustworthy or unsettling. That trained eye is exactly what separates a person who generates 200 mediocre images from a person who can look at 200 images and know instantly which one works and how to fix it. AI raises the value of taste while gutting the value of execution. Taste comes from reps. There is no shortcut, and now the shortcut everyone else is taking is the reason your reps are worth more.

Skills you outsource are skills you lose. Fewtrell makes a point in his essay that most people skim past: over-reliance on AI erodes the thinking it replaces. Creative muscles are muscles. If the machine does every rep for you, don't be shocked when you can't lift anything on your own. The designers who never learn foundations aren't becoming "AI-augmented creatives." They're becoming people who press a button and hope, with no ability to judge or repair what comes out.

Proof of human is becoming a luxury good. In the US, vinyl outsold CDs 46.8 million units to 29.5 million in 2025. Live Nation grew revenue 9% to US$25.2 billion in 2025, with a record 159 million people at its shows, in the age of infinite streaming. Handmade goods often command a premium over factory-made ones. The pattern is ancient and reliable: the moment something becomes infinitely reproducible, the scarce version becomes a status symbol. A hand-drawn identity, a real illustration style, design decisions that visibly came from a person, these are becoming the premium tier of visual work. In the Guardian piece, artist Rachel Maclean is already there. While everyone else waits for the models to stop generating eight-fingered hands, she's deliberately preserving the broken models because the errors are the interesting part. That's an artist using AI, not an artist replaced by it. The difference is that she has the craft to know which mistakes are beautiful.

The window is open right now

Herndon and Dryhurst offer one warning in that Guardian piece, and you should take it seriously. As the tools flatten technical skill, they predict the creative industry drifts toward a popularity contest, where "the most attractive kid in class" will have the tools to choose to be the most popular metal artist or crime thriller author. When everyone has the same generator, attention flows to whoever already has an audience.

There are only two defenses against that world. Be famous already. Or make work that is visibly, provably, unmistakably yours. Most of us don't get option one. Option two is learnable, and almost nobody is learning it right now, which is precisely why it's about to pay so well.

Think about the supply curve for a second. Plenty of juniors are skipping fundamentals and going straight to prompting. The pipeline of people who can genuinely draw, paint, and design from first principles is at risk of shrinking at the exact moment demand for distinctly human work is climbing. You don't need an economics degree to see what happens to the price of a skill when supply collapses and demand rises.

So here's the plan

If you're a creative, or you're raising one, or you're building a brand that needs to stand out in a feed full of synthetic sameness, do this:

• Learn the fundamentals with your hands. Drawing, typography, colour, composition. Analog first. The struggle is not an obstacle to the skill. The struggle is the skill.
• Use AI like an intern, never like an artist. Let it explore variations, mock up throwaways, handle grunt work. Never let it make the decisions that define how your work looks and feels.
• Show your process in public. Sketches, drafts, worksheets, time-lapses. Process is proof of human, and proof of human is the new trust signal.
• Develop a style a model can't average into existence. Weird, specific, personal. The middle is taken. The machine lives there now.
• Charge for judgement, not output. Output is free now. Knowing what's good is not.

What this means if you're buying creative work

This isn't only a question for artists. If you run a business, you're on the other side of the same trade, and the repricing hits your brand too.

When every competitor can generate polished visuals for free, polished stops being a signal. Customers learn to read machine-made imagery as "ad" and scroll past. The work that stops them is the work that visibly came from a person with a point of view: a real photo of your real team, an illustration style nobody else has, a campaign idea that could only have come from someone who understands your customer. We make the case for that shift in the authenticity premium.

So when you hire creative help, change the questions you ask:

• Can they show their thinking, not just their outputs? Sketches, references, rejected directions. A creative who can explain why one option beats another is selling you judgement, the scarce part.
• Do they have a recognisable style? If their portfolio looks like a generic image feed, your brand will too.
• How do they use AI? The right answer isn't "never" and it isn't "for everything". It's a clear line: AI for exploration and grunt work, humans for the decisions that define the look.
• Will the work still be yours in five years? A distinct visual identity compounds. A generated one can be matched by the next business that types a similar prompt.

Brands that invest in human-made, distinctive work now are buying the same thing the artists above are building: something that can't be averaged into existence. As we argue in why brand is the last moat, that's the one asset that gets more valuable as everything else gets easier to copy.

Art isn't dying. It's being repriced.

The camera didn't kill painting, it killed portrait-copying and gave us impressionism. Recorded music didn't kill performance, it made live shows the main event. AI won't kill art either. It will do what every disruptive technology has done: destroy the market for the average and multiply the reward for the exceptional.

Fewtrell is right. Art only dies when the artist dies. But careers die a lot easier than art does, and the careers that die in the next five years will belong to the people who let a text box do their seeing for them.

The pencil isn't obsolete. It just became the most contrarian, highest-leverage tool in the building. Pick it up.

And if you want a brand built by humans who did the reps, with the taste to know exactly when the machines are useful and when they're poison, that's literally what we do all day. Our photo and video team shoots real people, real products and real places, and uses the machines only where they make the human work better.

Frequently asked questions

Will AI replace artists and illustrators?

AI will replace a lot of average, on-demand creative output, because generating decent images is now nearly free. It won't replace artists with trained taste, a distinctive style and the judgement to know what works and why. Those skills become more valuable as generic output floods the market. The risk falls on creatives whose only value was producing something passable quickly.

Is it still worth learning to draw in the age of AI?

Yes, arguably more than ever. Drawing trains your eye for composition, light and form, which is exactly the judgement you need to direct AI tools well and to fix what they get wrong. As fewer people learn fundamentals, the people who have them become scarcer and easier to charge a premium for.

How should creatives use AI without losing their skills?

Treat it like an intern. Let it explore variations, mock up throwaway options and handle repetitive grunt work. Keep the decisions that define your work, such as style, composition and the final call, in your own hands, and keep practising fundamentals so you can still judge and repair what the machine produces.

Is AI-generated art bad for a brand?

Not automatically, but it carries risks. Generated imagery tends toward a generic, familiar look that makes brands blend together, and audiences increasingly recognise and distrust it. Use AI for concepting, mood boards and production tasks, and keep the visuals customers use to judge your brand, like your product, people and identity, human-made and distinctive.

Should you use AI-generated product photos in your ads? Not for anything that shows what a customer is actually buying. Use AI to speed up real creative, never to replace the truth of your product, because faked imagery is getting easier to detect, riskier to publish and more damaging to trust.

The company's Superintelligence Labs announced Muse Image in July 2026, a generative model built to let advertisers restyle product shots, pull still images out of video, and spin up endless creative variations without a photographer, a studio, or a single real photo. Meta says it uses "agentic visual reasoning and self-refinement" to understand a creative brief "the way a designer would." Early testers report better photorealism and better "product integrity." And more than 8 million advertisers are already using at least one of Meta's generative ad tools.

On paper, it sounds like a gift. Free photography. Infinite variations. No shoot days, no models, no location fees. Shoppers can even upload a photo of their own room and Meta AI restyles it with real products from business catalogues.

So here's the uncomfortable question nobody at Meta is going to answer for you: if generating a photorealistic product shot now costs nothing, what happens to a market where everyone's product shots cost nothing?

The answer is not "everyone wins." The answer is that the brands who lean on this hardest are quietly loading a gun and pointing it at their own reputation. Let me show you exactly how.

The seduction is real, and that's the problem

Understand why this is so tempting, because the pull is genuine. A proper product shoot can run four or five figures. AI generation runs a prompt. When you're a founder staring at a marketing budget, "make it look like we spent $10k" for free feels like a cheat code.

But a cheat code that everyone has isn't an advantage. It's the new baseline. And when everyone is generating the same glossy, over-lit, suspiciously-perfect imagery from the same handful of models, the entire category starts to look identical. Scroll your own feed for sixty seconds. You can already feel it, that faint plasticky sameness. The industry has a name for it now: AI slop. And audiences have developed an immune response to it.

Reason #1: AI images get caught, by machines and by people

The biggest myth about generative imagery is that it's undetectable. In 2026, it is anything but.

The machines are watching. A coalition of the largest tech and camera companies, Adobe, Microsoft, Google and Meta itself among them, has been baking provenance metadata directly into generated files through standards like C2PA / Content Credentials, alongside invisible watermarks like Google's SynthID. Meta already labels AI-generated content across Facebook and Instagram with "AI info" tags. In other words: the same platform selling you the generator is building the detector. Your "photorealistic" ad can arrive at the auction pre-flagged as synthetic.

And the humans are watching too. Consumers have trained their eyes fast. The tells are everywhere once you know them, hands with the wrong number of fingers, text that melts into gibberish, reflections that don't obey physics, skin with that waxy, poreless sheen, backgrounds that dissolve if you look too long. Your customer may not be able to articulate why an image feels off. They just feel it. And "off" is the last thing you want your brand to feel like at the exact moment someone is deciding whether to trust you with their money.

Reason #2: the consequences are getting expensive

Detection would be a footnote if it didn't come with a price tag. It does.

Regulators are closing in. The EU AI Act's transparency rules require AI-generated or manipulated content that could pass as authentic to be disclosed as such, and those obligations have applied since 2 August 2026, with some machine-readable marking duties following in December. In the US, the FTC has been explicit that there is no AI exemption from the laws on the books, so deceptive advertising made with AI is still deceptive advertising, the tool doesn't launder the lie. "We didn't take the photo, the AI did" is not a defence. It's a confession.

Trust is the real currency, and it's non-refundable. The moment a customer realises the "product" they fell in love with was a generated fiction, that the fabric doesn't drape like that, the colour isn't that colour, the room never existed, you haven't just lost a sale. You've taught them that everything you show them might be fake. That doubt doesn't stay contained to one ad. It infects your whole brand. In a world drowning in synthetic content, being the brand people can believe is worth more than being the brand with the most variations.

And "product integrity" is a marketing word, not a guarantee. Meta's early testers say Muse Image preserves your product better than before. Better is not the same as accurate. AI still hallucinates details, it invents a seam, changes a logo, adds a button that isn't there. When your ad shows something your product isn't, you've manufactured your own "expectation vs. reality" moment. Those go viral. Rarely in your favour.

Then there's the quiet cost nobody budgets for: returns, refunds and complaints. Every customer who receives something that doesn't match the ad is a return to process, a support ticket to answer and a review waiting to be written. Cheap imagery that inflates expectations doesn't save money. It moves the cost further down the line, where it's harder to see and more expensive to fix.

Reason #3: sameness is a death sentence for a brand

Here's the strategic core, and it's the part most people miss while they're busy being impressed by the technology.

A brand's entire job is to be distinct, to be instantly, unmistakably you. Generative models are trained on the average of everything that already exists. By definition, they pull toward the middle. Lean on them for your core visual identity and you will produce competent, convincing, and utterly forgettable work that looks like everyone else who typed a similar prompt. You will have automated your way into invisibility.

The scarce thing in 2026 is not more images. We are drowning in images. The scarce thing is proof, personality, and truth, and none of those can be prompted. It's the same argument we make about copy in why brand is the only thing left to buy when AI writes everyone's copy, and it applies to pictures even more.

The alternative: lean into what AI can't fake

The winning move isn't to reject technology and shoot everything on film like it's 1998. It's to be ruthless about where authenticity is non-negotiable, and to make that authenticity your unfair advantage while your competitors race each other to the bottom of the generated-content barrel.

That means:

• Real photography of your real product. The actual thing, the actual texture, in the actual light. This is now a differentiator, not a cost centre. It signals: we have nothing to hide. Our guide to product photography that makes people tap "Buy" covers what that looks like done well.
• Real faces and real proof. Genuine customers, genuine results, genuine founder-in-the-room content. Trust is transferred by humans, not by rendered ones.
• Signature visual identity. A distinct colour system, art direction, and photographic style that a model can't reproduce because it's yours, built deliberately, not averaged out of a training set.
• Provenance as a flex. As disclosure becomes law, "shot in-house, no AI" stops being a caption and becomes a positioning statement. Authenticity is about to be a marketable badge. Own it early. We explore that opening in the authenticity premium.

So where does AI belong?

Everywhere it doesn't touch the truth of your product or the trust of your audience.

Use it as the ferocious assistant it actually is: rapid concepting and mood-boarding before a real shoot, background cleanup and retouching on genuine photos, resizing and reformatting a single hero asset into fifty placements, storyboarding, and testing layout ideas fast and cheap. That's leverage. That's AI doing the tedious 80% so your team can pour real craft into the 20% that customers actually judge you on.

The line is simple: use AI to accelerate real creative, never to replace it.

A five-step AI image policy for your brand

Most businesses don't have a rule about this, which means whoever is building ads that day decides. Write it down. It takes an hour.

1. Define the "truth zone." List every image type that shows what a customer receives: product shots, colours, materials, sizes, finished work, before-and-after results, your team and your premises. These must be real photography, full stop.
2. Define the "assist zone." List where AI is welcome: concept boards, storyboards, background cleanup on real photos, resizing, layout tests, abstract graphics that don't depict a product or a person.
3. Set a disclosure rule. Decide how you'll label AI-assisted imagery where required, and follow the platforms' labelling requirements rather than trying to dodge them.
4. Add a truth check before publishing. Someone who knows the product compares every ad image against the real item. Wrong colour, extra detail or missing feature means it doesn't go live.
5. Build a real image library. Plan shoots so you have a steady supply of genuine assets, which removes the temptation to fake a gap on deadline day.

A worked example: the furniture store and the empty room

Here's a hypothetical. Say you run an online furniture store in Melbourne. A generative tool lets you drop your sofas into gorgeous, sun-drenched rooms for next to nothing. The ads look incredible. Clicks jump.

Then the orders arrive. The fabric in the ad looked richer than it is. The tool made the sofa look slightly deeper than its real dimensions. Customers feel misled, returns climb, and a few leave reviews with side-by-side photos of the ad and the delivery. Those reviews outlive the campaign.

Now picture the alternative. You shoot your sofas properly in two real rooms with honest light and close-ups of the fabric. You use AI to extend backgrounds for different ad sizes and to test layouts before the shoot. Your ads look a little less like a fantasy and a lot more like what turns up at the door. Fewer surprises, fewer returns, better reviews. This is an illustration, not a real case, but it's the trade every product brand is facing right now.

Real beats rendered

The brands that win the next few years won't be the ones who generated the most. They'll be the ones who stayed unmistakably, verifiably real while everyone else blurred into the same synthetic sludge. Artists face the same fork, as we argue in AI won't kill art, but it will bury artists who never learned to draw.

Meta just handed 8 million advertisers the same button. The ones who mash it hardest will look exactly like each other. The ones who use it with a strategy, and protect the parts of their brand that have to be true, will be the ones people actually remember, believe, and buy from. That's the work our photo and video team does: real imagery with a signature look, built to be believed.

Faking it has never been cheaper. Which is exactly why being real has never been more valuable.

Frequently asked questions

Is it legal to use AI-generated images in ads?

Using AI imagery isn't illegal in itself, but misleading advertising is, however the image was made. Regulators like the US FTC treat AI-created deceptive ads as deceptive ads, and the EU AI Act adds disclosure obligations for AI-generated and manipulated media. If an image misrepresents your product, the tool you used is no excuse.

Can people tell when a product photo is AI-generated?

Often, yes. Platforms and standards like C2PA Content Credentials, Google's SynthID and Meta's "AI Info" labels can flag synthetic images, and many consumers spot the visual tells, from warped text to waxy skin and impossible reflections. Even when people can't say why, an image that feels slightly off erodes trust at the moment of purchase.

What should I use AI for in my visual marketing?

Use it where it doesn't touch the truth of your product: mood boards and concepts before a shoot, storyboards, background cleanup and retouching on real photos, resizing one hero image into many placements, and testing layouts. Keep product shots, colours, results, people and premises as genuine photography.

Are AI product photos cheaper than a real photoshoot?

Upfront, usually. Over time, often not. Generated images that misrepresent a product can drive returns, refunds, complaints and bad reviews, and they make your brand look like every competitor using the same tools. A well-planned real shoot produces assets you can use for months and builds the trust that keeps customers buying.

So what is AI saying about you? Whatever the web gives it to work with. Assistants build their answer from your website, your reviews, directories and anything else written about you. If that material is clear, consistent and full of proof, the verdict is flattering. If it's thin or contradictory, the verdict is vague, wrong, or about somebody else.

AI is the new word of mouth

Answer engines don't hand back ten blue links for the buyer to judge. They synthesise one authoritative-sounding verdict. If that verdict is vague, outdated, or built from a competitor's marketing, you have already lost the deal before a human ever saw your homepage.

Think about how word of mouth used to work. A friend says "use these guys, they were great." You trust it because it's short, specific and comes from someone with no reason to lie. An AI answer feels exactly like that. It doesn't read like an ad. It reads like advice. Which is why buyers take it at face value, and why a lukewarm description from an assistant can quietly kill a sale you never knew you were in.

The difference is scale. A friend recommends you to one person. An assistant answers the same question for everyone who asks it, all day, every day. Whatever it has decided about you gets repeated to your whole market.

You can't be in the room when AI describes your brand. So you'd better shape what it has to work with.

Where the AI's opinion of you comes from

Assistants aren't forming opinions out of thin air. They draw on a few broad sources, and knowing them tells you exactly where to work.

• Your own website. What you say you do, where, and for whom. If it's buried in slogans, the model has little to lift.
• Your business listings. Google Business Profile, industry directories, marketplaces and social bios. These are often where the contradictions live.
• Reviews. What customers say about you, in their words, on sites the models can read. We cover why this matters so much in why AI reads your reviews before it recommends you.
• Independent mentions. Articles, roundups, podcasts, forum threads and comparison pages written by people who aren't you.
• Live search results. Many assistants, ChatGPT included, now search the web in real time before answering, so what ranks for your name and category feeds straight into the answer.

Notice that only one of those five is fully under your control. That's the uncomfortable part. It's also the opportunity, because most of your competitors are only working on the one they control.

How to become the answer, not the afterthought

• Be unambiguous about who you are: Clear, consistent descriptions of what you do, for whom, and why you're different, on your site and everywhere else the models read. Confusion in, confusion out.
• Get cited by sources AI trusts: Reviews, credible mentions, guest articles, and directories are the raw material these systems ingest. Earn presence where the training data lives.
• Structure your site for machines: Schema markup, clean FAQs, and direct answers up top turn your pages into things an AI can quote confidently.
• Feed it real proof: Named results, specific numbers, and genuine expertise give the model concrete, favourable things to repeat about you.

Each of those deserves a closer look, because "be clear" is easy to agree with and hard to actually do.

Clarity means writing one plain sentence that describes your business and using it everywhere. "We build custom timber decks for homes on the Sunshine Coast" beats "We create outdoor living experiences" every single time, because a machine can repeat the first and can do nothing with the second. The principle behind this is covered in becoming an entity AI trusts.

Citations mean getting described by other people. Ask happy customers for reviews that mention what you did for them. Pitch a useful guest article to an industry publication. Make sure you're listed accurately in the directories your buyers actually use.

Structure means pages that answer questions directly, with the answer in the first line and the proof underneath. Add FAQ sections using the questions buyers really ask. Mark them up.

Proof means specifics. Years in business, service areas, turnaround times, guarantees, the types of clients you work with. Vague praise gives the model nothing. Concrete detail gives it a sentence to say about you.

A worked example: the builder the AI got wrong

Say you run a renovation company in Geelong. You moved out of Melbourne three years ago, stopped doing new builds and now specialise in kitchens and bathrooms. Business is good. Then a prospect mentions on a call that ChatGPT described you as "a Melbourne builder offering new homes and extensions."

Where did that come from? You go looking. Your old website copy is still on a cached directory listing. A local business roundup from years ago still lists your old address. Your Google profile is right, but your LinkedIn says "residential builder." Your own homepage says "transforming spaces" and never actually says kitchens and bathrooms in Geelong.

So the model did its best with what it found, and its best was wrong.

The fix is methodical. Rewrite your homepage opening to say exactly what you do and where. Update every listing and profile to match, word for word. Ask your last twenty happy clients for a review that mentions their kitchen or bathroom. Publish a short guide answering the questions Geelong homeowners ask about bathroom renovations. Then re-run the same prompt every month and watch the description move toward the truth.

Nothing about that is technical wizardry. It's housekeeping that nobody did, because nobody was checking what the machines were saying.

Audit yourself today

Open your favourite AI assistant and ask it about your company, your category, and your top competitor. What it says, and whether it mentions you at all, is the most important brand audit you'll run this year. If you don't like the answer, that's not a problem to ignore. It's the exact work that decides who wins the next decade of search.

To make the audit useful rather than just alarming, run it properly:

1. Use several assistants. ChatGPT, Gemini, Perplexity and Google's AI results can give different answers because they draw on different sources.
2. Ask three kinds of question. Your brand by name ("What do people say about [your business]?"), your category ("Who's the best [service] in [city]?") and a comparison ("[You] or [competitor], which should I choose?").
3. Write down what comes back. Are you mentioned? Is the description accurate? Which sources are cited? What does it say about your competitor that it doesn't say about you?
4. List every error and gap. Wrong location, old services, missing specialties, no mention at all.
5. Trace each one to a source. Most errors come from a specific listing, page or article. Fix the source, not just your homepage.
6. Repeat monthly. Save your prompts and run the same ones each month so you can see real movement over time.

The mistakes that make the answer worse

• Fixing only your own site. If the directories and old articles still say the wrong thing, the model still has contradictions to deal with.
• Writing for keywords instead of facts. Pages stuffed with search phrases but light on plain statements give assistants nothing solid to repeat. See the three edits that get your business quoted for what to do instead.
• Buying fake reviews. It's dishonest, the ACCC says it is against the law and review platforms remove them. It also teaches the model nothing true about you.
• Checking once and forgetting. The answers change as the web changes. A one-off audit is a snapshot, not a strategy.

What to do first this week

You don't need a six-month project plan to start. You need five focused hours spread across the week.

1. Monday: run the audit. Ten prompts across three assistants. Paste the answers into one document so you can see the whole picture at once.
2. Tuesday: write your one-sentence description. What you do, for whom, and where. Plain words. Read it aloud. If a stranger couldn't repeat it back, simplify it.
3. Wednesday: fix your top five listings. Google Business Profile, LinkedIn, Facebook and your two most important industry directories. Same name, same description, same service area, same contact details.
4. Thursday: rewrite your homepage opening. Put the one-sentence description at the top, followed by the proof: how long you've been doing it, who for, and what customers get.
5. Friday: ask for reviews. Email ten recent happy customers. Ask them to mention the specific job you did and where. Specific reviews give the models specific things to say.

That's one week. It won't fix everything, but it removes the most common causes of a bad AI answer, and it gives you a baseline to measure against next month.

This is brand management now

For decades, brand management meant controlling your logo, your ads and your website. Now a big slice of your reputation is written by a machine summarising everything else. You can't control that summary directly. You can control the raw material, and that turns out to be most of the game.

The work is steady rather than dramatic: clear descriptions, consistent listings, real reviews, useful pages and regular checks. It's exactly what our digital marketing team does for businesses who want to show up well when their buyers ask a machine for advice. Run the audit today. If you like what you hear, protect it. If you don't, you now know what to fix.

Frequently asked questions

How do I find out what ChatGPT says about my business?

Ask it directly. Try your business name, your category plus your city, and a comparison with a competitor. Repeat the same prompts in Gemini, Perplexity and Google's AI results, because each can answer differently. Note whether you're mentioned, how accurately you're described and which sources are cited. Save the prompts and re-run them monthly to track whether the picture improves.

Why does AI give wrong information about my company?

AI assistants build answers from what they can find online, including old directory listings, outdated articles and inconsistent profiles. If your details differ across sources, the model may pick the wrong version or blend several together. The fix is to trace each error to its source, correct it, and make your business description consistent across your website, listings and social profiles.

Can I pay to be recommended by AI assistants?

Assistants like ChatGPT and Perplexity form their organic recommendations from the content they find, not from payments. OpenAI says ads in ChatGPT do not influence its answers. The reliable way in is earning it: clear information on your own site, consistent listings, genuine reviews, independent mentions and pages that answer buyer questions directly. Be wary of anyone promising guaranteed AI recommendations for a fee, because nobody controls what the models say.

How long does it take to change what AI says about you?

It varies. Assistants that search the web live can reflect updated pages and listings fairly quickly once search engines recrawl them. Answers that lean on a model's built-in training data change more slowly, often only when the model is updated. That's why consistent, ongoing work beats a one-off fix. Track it monthly with the same prompts so you can see progress.

So how do you make AI trust your brand? You make it easy for the machine to be certain about you. Describe your business identically everywhere, get listed in the structured sources engines use to define entities, earn mentions from sites the model already trusts, and cover your topic so thoroughly that the connection between your name and your subject becomes obvious. Certainty is what gets you named.

Most businesses have never thought about this. They have a website, a few profiles set up years ago by different people, a directory listing with an old phone number, and a LinkedIn page that describes a service they stopped offering. To a human, that is mild untidiness. To a machine trying to decide whether you are one business or three, it is a reason to recommend someone else.

What an entity actually is (and why keywords are not enough)

A keyword is a string of letters. An entity is a thing. "Apple" the keyword could mean fruit or a company. Apple the entity is a specific organisation with a founder, a headquarters, products and a history, and every fact about it is connected in a graph of other things.

When an AI assistant answers "who is the best accountant for tradies in Newcastle," it is not matching those words against your homepage. It is looking for entities it knows are accountants, that are connected to Newcastle, that are associated with tradespeople, and that have enough positive corroboration to be worth naming. If the machine has never confidently connected your name to those attributes, you are not a candidate, however good your service is.

That is the shift. Keywords made you findable. Entity status decides whether you get recommended. We unpack the wider mechanics in AEO, GEO and SEO explained, but the practical work starts here.

Be the same thing everywhere

Consistency is how a model connects the dots that you and "that company" are one entity. Use identical business names, descriptions, and contact details across your site, profiles, and every listing. Contradictions fragment your identity and dilute your authority.

This sounds basic. It is also where most businesses quietly fail. Here is what "identical" really means:

• One exact business name. Not "Smith Plumbing" on the site, "Smith Plumbing Services Pty Ltd" on Google and "Smith's Plumbing" on Facebook. Pick the version customers use and use it everywhere.
• One core description. Write a two-sentence statement of who you are, what you do and who you do it for. Use it, word for word or very close, on every profile that has an "about" field.
• One set of contact details. Same address format, same phone number, same website URL. Old numbers and old addresses floating around are contradictions the model has to resolve.
• One list of services. If your site says you do five things and your directory listing says eight, the machine has to guess which is true. It usually guesses by trusting you less.

Get into the sources that define entities

• Knowledge bases: A complete Google Business Profile, LinkedIn, Crunchbase, and Wikidata presence gives engines structured facts to anchor to.
• Credible mentions: Being referenced by publications and sites the model already trusts transfers a slice of that trust to you.
• Reviews and social proof: Volume and consistency of what others say about you shapes the sentiment AI repeats.

Treat these as three layers. The knowledge bases tell the machine what you are. The credible mentions tell it that other trustworthy sources agree. The reviews tell it what you are like to deal with. Miss one and the picture is incomplete.

A word on Wikidata: it is a structured, community-edited database with its own notability standards. Do not create an entry that stretches the truth or reads like an ad, because it will be removed and may damage your credibility with the editors. Only pursue it when you have genuine independent coverage to point to.

Reviews deserve their own attention too, because they are one of the few entity signals you can grow every single week. We go deeper on that in why AI reads your reviews before it recommends you.

Keywords made you findable. Becoming an entity makes you trustworthy. Only trusted entities make it into the answer.

Own your topic, not just your name

Trust compounds when you cover a subject completely. Interlinked clusters of content, a pillar page plus every supporting angle, signal that you don't just mention a topic, you own it. Add Organization, Person, and sameAs schema so the connections are spelled out in machine-readable form.

Here is what that looks like in practice:

• A pillar page for each core service that answers the big question fully: what it is, who it is for, how it works, what it costs, what to watch out for.
• Supporting articles for every narrower question a buyer asks along the way, each linking back to the pillar and to each other where it helps the reader.
• Organization schema on your site that states your official name, logo, contact details and founding information.
• Person schema for founders and key experts, because people are entities too, and a named expert attached to your business adds credibility.
• sameAs links inside that schema pointing to your official profiles, so the machine does not have to guess which LinkedIn page or which directory listing is really you.

Schema does not replace reputation. It removes ambiguity. It turns "probably the same business" into "definitely the same business," and that certainty is exactly what you are trying to build.

A step-by-step entity cleanup you can start this week

1. Write your canonical facts. Official name, two-sentence description, services, service area, address, phone, website, founders. One document. This is the source of truth.
2. List every place you exist online. Search your business name, your phone number and your old addresses. Note every profile, directory and mention you find.
3. Fix the contradictions. Update every profile you control to match the canonical facts. Request corrections on the ones you do not control.
4. Complete the big profiles properly. Google Business Profile with categories, services, photos and hours. LinkedIn with the same description. Crunchbase if you are a company that would reasonably be listed there.
5. Add schema to your site. Organization, Person where relevant, and sameAs pointing to the profiles you just cleaned up.
6. Map your topic cluster. Pick your main service, write the pillar, list the ten questions buyers ask, and schedule the supporting pieces.
7. Start earning mentions. Industry associations, supplier pages, local business roundups, podcast guest spots, genuinely useful answers in community forums. Independent sources are what turn your claims into corroborated facts.

A worked example: the physio clinic with three identities

Say you run a hypothetical physiotherapy clinic in Geelong. You rebranded two years ago. The website uses the new name. Google Business Profile still carries the old one. A health directory lists your former address. Your LinkedIn page describes you as a "wellness studio," which was a phase.

Ask an assistant about sports physios in Geelong and you are nowhere. Not because the model thinks you are bad. Because it holds fragments of three businesses and cannot confidently connect any of them to "sports physiotherapy in Geelong."

Now run the cleanup. One name everywhere. One description that says exactly what you do: sports injury physiotherapy in Geelong for runners, footballers and weekend athletes. Old directory entries corrected. Organization schema with sameAs links. A pillar page on sports injury rehab and supporting articles on the specific injuries your patients ask about. A handful of mentions from the local running club and a sports association.

Nothing about the clinic's quality changed. What changed is that the machine can now state facts about you with confidence. That is usually the difference between being ignored and being named.

Common mistakes that keep you invisible

• Leaving old profiles alive. A forgotten listing with outdated details keeps contradicting you. Update it or remove it.
• Describing yourself differently for every platform. Clever variations feel fresh to you and read as confusion to a machine.
• Relying on your own site alone. A brand that only talks about itself on its own domain has no corroboration. Trust comes from others.
• Chasing volume over coverage. Fifty loosely related blog posts do less than ten that fully cover one subject and link together properly.
• Treating schema as a magic switch. Markup clarifies. It does not create authority that the rest of the web does not support.

Run the audit

Ask an AI assistant to describe your brand and your category. If it's vague, wrong, or silent about you, that's your entity gap laid bare, and closing it is the highest-leverage marketing work you can do this year.

Make it systematic. Ask ChatGPT, Gemini, Perplexity and Google's AI answers the same set of questions: "What is [your business]?", "Who are the best [your service] providers in [your area]?", and "What is [your business] known for?" Record the answers in a spreadsheet. Repeat monthly. Our guide to what AI is telling your customers about you walks through how to read the results.

Look for three things. Is the description accurate? Are you named in category questions, or only when asked directly? Which competitors appear instead, and what sources back them up? Those answers become your to-do list.

Make the machine certain about you

Entity work is not glamorous. It is spreadsheets, profile fields, schema and slow, steady reputation building. That is exactly why it works. Your competitors are still tweaking meta descriptions while the systems deciding who gets recommended are reading the whole web for consistency and corroboration.

If you would rather have it done properly, our digital marketing team runs entity audits, cleanups, schema and topic clusters as one joined-up project, so the machine sees one clear, trustworthy brand instead of a pile of fragments.

Frequently asked questions

What does it mean to be an entity in search?

An entity is a distinct, identifiable thing, such as a person, business, place or concept, that search engines and AI models store with defined attributes and relationships. Being an entity means the machine knows who you are, what you do, where you operate and how you connect to other things. That recognition is what allows an assistant to confidently recommend you.

How do I get ChatGPT to mention my business?

There is no submission form. You earn mentions by making your business easy to understand and well corroborated. Keep your name and description consistent everywhere, complete major profiles like Google Business Profile and LinkedIn, earn mentions on credible independent sites, collect detailed reviews, and publish thorough content on your core topic. Then check regularly what assistants say about you.

Does schema markup help AI understand my brand?

Yes, as a clarity tool. Organization and Person schema state your key facts in machine-readable form, and sameAs links connect your site to your official profiles so there is no doubt they belong to the same business. Schema will not create authority on its own, but it removes ambiguity, which makes the trust you have earned elsewhere easier to recognise.

How long does it take to build entity authority?

Cleanup work, such as fixing inconsistent listings and adding schema, can be done in weeks. Earning independent mentions, reviews and topic authority is ongoing and compounds over months. AI models also update on different schedules, so some changes show up faster in assistants that search the live web than in those relying more on training data. Measure monthly and stay consistent.

The short answer: to get quoted by ChatGPT, Perplexity or Google's AI Overviews, put the direct answer first, break the page into self-contained blocks under question-style headings, and publish things only you can say, like your own numbers, comparisons and first-hand experience. Clear, specific, liftable sentences get cited. Vague, meandering ones get skipped.

Most businesses already know their subject better than the pages that get quoted instead of them. They just write it in a shape a machine cannot use. Let's fix that.

Why AI quotes some pages and ignores others

An answer engine is not reading your page the way a person does. It is looking for passages that resolve a question cleanly, that make sense when lifted out of context, and that come from a source it has reason to trust. Then it stitches those passages into an answer and, if you are lucky, names you.

That changes what "good content" means. A beautifully written essay that builds slowly to its point can be a joy to read and almost useless to a model. The point is buried, the paragraphs lean on each other, and no single sentence stands up on its own.

Meanwhile a plainer page with a crisp first sentence, clear headings and one idea per section gets pulled apart and quoted all over the place. Same facts. Different shape. Different result. If you want the bigger picture on how this fits alongside traditional SEO, our breakdown of AEO, GEO and SEO covers the whole alphabet.

Lead with the answer

Ditch the slow windup. State the conclusion in the first sentence or two, then support it underneath. Machines grab the first clear statement that resolves the query, so bury your answer three paragraphs down and you hand the citation to someone else.

This is the same rule newspapers have used for a century: the inverted pyramid. The most important thing goes first, the detail follows, the background comes last. It works for machines for the same reason it works for busy humans. Nobody has to hunt.

Here is the test. Read only the first two sentences under any heading on your page. Do they answer the question the heading asks? If they are scene-setting, history or "great question!" filler, cut them or move them down.

• Weak opener: "When it comes to choosing how to bill clients, there are many factors every business needs to consider, and every situation is different."
• Strong opener: "For most small service businesses, a monthly retainer beats hourly billing if you want predictable cash flow, while hourly billing suits one-off jobs where the scope is unclear."

The second version can be lifted straight into an answer. The first cannot.

One idea per chunk, headings as questions

Break content into tight, self-contained blocks under headings phrased the way people actually ask, "How much does X cost?", "X vs Y: which is better?" Each block should stand alone as a complete answer a model can lift cleanly.

"Self-contained" is the key word. If a paragraph starts with "As mentioned above" or "This is why", it depends on something the model may not grab. Name the subject again. Repeat the key noun instead of relying on "it" and "this". It feels slightly redundant to you. To a model extracting one block at a time, it is the difference between usable and useless.

Where do the questions come from? Your inbox. Your sales calls. The questions customers ask before they buy are the questions they type into an assistant. Keep a running list and turn the best ones into headings.

The formats AI loves most

• Original statistics: A specific, sourced number is the single most citable thing you can publish. Run a survey, share your own data, coin a stat.
• Comparisons and "vs" tables: Buyers ask engines to compare options constantly, give them the clean side-by-side to quote.
• Numbered steps and lists: Structured, scannable, and effortless for a machine to reproduce in an answer.
• Plain-English definitions: Clear "X is…" statements get pulled straight into explainer answers.

A note on original statistics, because this is where people get tempted to cut corners. "Coin a stat" means publish a real number from your own work, not make one up. Average turnaround time across your jobs this year. The most common fault you found in the last hundred inspections. What your customers spend on average before they come to you. Small, honest, specific data from your own business beats a recycled industry figure every time, because nobody else can publish it.

Comparisons work because they match how people shop. Nobody asks an assistant "tell me about heat pumps". They ask "heat pump or gas hot water, which is cheaper to run?" If your page answers that directly, with the trade-offs laid out honestly, you have given the model exactly what it needs.

Write for the human, structure for the machine. The pages that do both get read and quoted.

How to rewrite an existing page for AI citation

You do not need to start from scratch. Most businesses have pages with the right knowledge in the wrong shape. Here is the rewrite process, step by step.

1. Pick one page that should already be winning. Usually a service page or your most useful guide. Start where the knowledge is strongest.
2. List the questions it should answer. Five to ten, in customer language. Check them against real enquiries.
3. Turn the best ones into headings. Phrase them as questions or as clear statements of what the section resolves.
4. Rewrite the first sentence of every section as the answer. Then support it. Then add nuance.
5. Break up any paragraph that carries two ideas. One idea, one block.
6. Add one thing only you can say. A number from your own work, a comparison from real experience, a mistake you see constantly.
7. Add a short FAQ. Direct questions, answers of a few sentences, each one standing alone.
8. Put a real name on it. Author, role, experience. More on that below.

For the three highest-leverage edits on their own, see three edits that get your business quoted by ChatGPT.

A worked example: the plumber's hot water page

Say you run a plumbing business in Perth with a "Hot Water Systems" page. Right now it opens with two paragraphs about your family history, then a list of brands you install, then "call us for a quote". It has everything a model would want and none of it in usable form.

The rewrite opens with a direct answer: which hot water system suits most households in your area, and why, in two sentences. Then come headings built from the questions your office actually fields: "Heat pump vs gas vs electric: which is cheapest to run?", "How long does a hot water system last?", "What does hot water replacement cost?", "Signs your hot water system is about to fail". Under the cost heading you give honest ranges from your own jobs and explain what pushes a job to the top of the range.

Then you add the thing only you can say: the most common reason you see systems fail early in local conditions, from years of call-outs. Finally, the page is signed by the licensed plumber who wrote it, with their licence type and years in the trade.

Nothing about the business changed. The page now contains a dozen passages a model can lift cleanly, several of which exist nowhere else online. That is a hypothetical, but it is exactly the shape of the change.

Sign your work

Named authors with real credentials, first-hand experience, and genuine data signal a trust that thin, AI-spun content can't fake. The engines are actively hunting for that signal. Give them a reason to believe you're the source worth naming.

In practice that means a real author name on every article, a short bio stating what qualifies them, and an author page that links to their other work and their profiles elsewhere. It means writing "we tested", "on our last job", "the mistake we see most", because first-hand experience is the thing generated filler cannot fake. And it means your business details are consistent everywhere, so the model can connect the page to a real organisation. We go deeper on that in how to make AI trust your brand.

The mistakes that stop you getting quoted

• Writing for word count. Padding dilutes the good passages. A model has to dig through more noise to find the answer, and the answer often gets lost.
• Hiding answers behind a form or a call. "Contact us for pricing" gives the model nothing, so it quotes the competitor who published ranges.
• Clever headings. "The heat is on" tells a machine nothing. "How long does a hot water system last?" tells it everything.
• Mass-producing AI copy. If your page says what a model would say anyway, there is no reason to cite you. Original experience is the whole point.
• Burying key content in images or PDFs. If the answer lives in a graphic, many systems cannot read it. Put the text on the page.

How to tell if it is working

Measurement here is still rough, so combine a few signals. Ask the main assistants the questions your pages answer and note whether you are named or linked. Do it monthly and keep a simple log. Watch referral traffic from AI tools in your analytics. Watch branded search in Search Console, because people who see you recommended often search your name next. And ask new enquiries how they found you, adding "an AI assistant" as an option.

None of these is perfect on its own. Together they show a trend, and the trend is what matters.

Content that earns the citation

The businesses that win AI search will not be the ones publishing the most. They will be the ones whose pages are easiest to quote and hardest to replace: direct answers, clean structure, real numbers, real names.

That is the work we do in our digital marketing and AI search service at LIQID NOISE. We restructure the pages you already have, build the ones you are missing, and make sure the expertise inside your business actually shows up where your customers are asking. You already know the answers. We make sure the machines can find them.

Frequently asked questions

How do I get my website quoted by ChatGPT?

Make your pages easy to extract and worth citing. Put a direct answer in the first sentence under each heading, keep one idea per section, and phrase headings like real questions. Add something original, such as data from your own work or first-hand comparisons. Sign content with a named, qualified author. Clear, specific, self-contained passages from a trustworthy source are what answer engines lift.

What content format do AI answer engines prefer?

Formats that can be lifted cleanly into an answer. Plain-English definitions, numbered steps, honest comparisons between options, short FAQs and original statistics all work well. The common thread is that each block makes sense on its own without the surrounding page. Long, winding paragraphs that build slowly to a point are the hardest format for a model to quote.

Does writing for AI search hurt readability for humans?

No, it usually improves it. Leading with the answer, using clear headings and keeping one idea per section is simply good writing for busy readers. People skim before they read, and a page structured for extraction is also easy to skim. The only thing you lose is the slow windup, and very few readers will miss it.

Should every article have an FAQ section for AI search?

Most informational pages benefit from one. A short FAQ gives answer engines ready-made question and answer pairs, and it catches the follow-up questions readers have after the main content. Keep answers direct and self-contained, use the wording customers actually search with, and avoid repeating the body word for word. A few strong questions beat a long list of thin ones.

The direct answer: to get cited by ChatGPT, Gemini and Perplexity, you need to be clearly understood (a consistent description of who you are), easy to quote (pages that lead with plain, self-contained answers), talked about in the places these tools read (reviews, forums, video, publications), marked up with structured data, and kept current. That's generative engine optimisation, or GEO. None of it is a trick. All of it compounds.

Most businesses are doing none of it, which is exactly why the ones who start now get an unfair head start.

What GEO actually is, and why it's different from SEO

Classic SEO was about ranking a page in a list of ten blue links. The searcher clicked, landed on your site and made up their own mind. GEO is about being the source an AI system draws from when it writes a single answer. There's no list to climb. You're either in the answer, named and recommended, or you're not there at all.

That changes what "good content" means. A search engine could rank a long, meandering page because the searcher would scroll to find the bit they needed. A language model building an answer wants a clean, confident statement it can lift and attribute. It also cross-checks. If your website says one thing and the rest of the web says something else, the model hedges, and hedging usually means leaving you out.

SEO still matters, a lot. Many AI tools pull from search results as part of how they find sources. But GEO adds new requirements on top. If the alphabet soup is confusing, our plain-English guide to AEO, GEO and SEO sorts out which is which.

1. Nail down your entity

Before AI can recommend you, it has to understand exactly what you are. Describe your business, your category, and your differentiators the same clear way everywhere online. Ambiguity gets you left out of the answer entirely.

In practice, write one "entity statement" and use it everywhere. Something like: "[Business name] is a [category] in [location] that helps [who] with [what], known for [differentiator]." Then make sure that same description, word for word where possible, appears on:

• Your homepage and About page
• Your Google Business Profile
• Your LinkedIn company page and founder profiles
• Industry directories and association listings
• Any podcast bios, speaker bios and guest post author boxes

The goal is that anything reading about you, human or machine, comes away with the same clear picture. We go deeper on this in how to become an entity AI trusts.

2. Publish extractable, quotable content

Models lift clean, self-contained statements. Lead every page with a direct answer, define terms plainly, and pack in specifics, numbers, steps, comparisons, that a machine can quote without needing the surrounding paragraph.

Here's a quick test. Take any paragraph on your site and paste it on its own into a blank document. Does it still make sense? Does it answer a question without relying on "as mentioned above" or "this"? If yes, it's quotable. If not, it's filler to a machine.

The patterns that get lifted most often:

• Question-shaped headings followed immediately by a one or two sentence answer.
• Definitions that start with "X is..." and finish in a single sentence.
• Numbered steps for anything procedural.
• Comparisons that spell out the difference between two options in plain terms.
• Specifics you actually own, like your pricing ranges, turnaround times, service areas and process, stated plainly instead of hidden behind "contact us."

For three concrete rewrites you can make to existing pages, read three edits that get your business quoted by ChatGPT.

3. Earn mentions where AI actually reads

Generative engines pull from the wider web, not just your site. Reviews, Reddit threads, YouTube, industry publications, and roundup articles are prime retrieval sources. Being talked about in those places is often worth more than another blog post on your own domain.

On your website you control the message. In the AI answer, the web's consensus about you is the message. Go shape it.

Shaping consensus doesn't mean astroturfing. Fake reviews and planted forum posts are a fast way to get caught and lose trust with people and platforms alike. It means doing the unglamorous work of showing up:

• Build a steady flow of genuine reviews on the platforms that matter in your industry, and reply to them. Detailed reviews that mention specific services give AI concrete things to repeat. We explain why in AI reads your reviews before it recommends you.
• Pitch yourself for "best of" roundups and comparison articles in your niche. These are exactly the pages AI tools consult when someone asks "who's the best X for Y?"
• Answer questions in public on Reddit, industry forums and Q&A communities, as yourself, adding real value, without a hard sell.
• Publish on YouTube. Explain your expertise on camera with a clear title and description. Video platforms are part of the web AI draws on.
• Get quoted in trade publications and local media. A named expert quote in a trusted outlet carries weight.

4. Mark everything up

Schema for your organisation, products, FAQs, and authors turns your pages into structured, machine-readable facts. Clean data is easier to trust and easier to cite.

The minimum set for most businesses:

1. Organization or LocalBusiness schema on your homepage, with your name, logo, address, contact details and links to your official profiles.
2. Service or Product schema on each service or product page, describing what it is and who it's for.
3. FAQPage schema on pages with a genuine FAQ section, where the questions and answers are visible on the page too.
4. Article and author markup on blog posts, connecting each piece to a real person with real credentials.
5. Review markup where your platform supports it and the reviews are genuine.

Then test it. Google's Rich Results Test and the Schema.org validator will tell you whether the markup reads correctly. Broken schema is worse than none, because it sends mixed signals.

5. Stay fresh and consistent

• Update your cornerstone content so it reflects the current year, stale pages get quietly dropped.
• Keep your facts identical across every profile and mention, contradictions erode the confidence a model needs to name you.
• Feed the engines proof, named results, real data, and credentials give the AI concrete, favourable things to repeat.

Put a date in the calendar every quarter to review your ten most important pages. Update the facts, prices and examples, refresh the "last updated" date only when you've made a real change, and check every external profile still matches. When you change your pricing, your address or your service list, change it everywhere on the same day. Contradictions are how models lose confidence in you.

How to tell if GEO is working

You can't open a dashboard and see your "AI ranking." But you can measure it well enough to make decisions.

• Run a monthly prompt check. Write down ten questions your ideal customer would ask an AI tool, like "best [service] in [city] for [situation]." Ask them in ChatGPT, Gemini and Perplexity each month. Record whether you're mentioned, how you're described, and which sources get cited.
• Watch referral traffic from AI tools in your analytics. Visits from chatgpt.com, perplexity.ai and similar sources show up as referrals, and they're worth segmenting on their own.
• Ask new leads how they found you. Add "AI assistant (ChatGPT etc.)" as an option on your enquiry form. Plenty of people will tell you straight.
• Track which sources get cited for your category. If one directory or publication keeps showing up, that's where you need to be listed.

A worked example: the physio clinic

Say you run a physiotherapy clinic in Perth. You ask ChatGPT "best physio in Perth for running injuries" and you're not mentioned. Two competitors are, and the answer cites a local "best of" article and a Reddit thread.

Here's a 90-day plan built on the playbook. Month one: write your entity statement ("a Perth physiotherapy clinic specialising in running and sports injuries") and push it across your site, Google Business Profile and directories. Rewrite your running injuries page to open with a direct answer, list the conditions you treat, explain your process step by step and add a real FAQ with schema. Month two: set up a simple review request after a patient's final session, asking them to mention what they came in for. Contact the author of that local roundup with a short, useful pitch. Month three: record five short videos answering common runner questions and publish them with clear titles.

Then rerun the prompt check. You might not be in the answer yet. But you've given every AI tool consistent, quotable, well-supported reasons to include you, and that's the part you control. (This scenario is hypothetical, but the steps are exactly what we'd prioritise.)

Where to start this week

Don't try to do all five at once. Start here: write your entity statement, fix any contradictions across your top five profiles, and rewrite the opening of your most important service page so it answers the core question in two sentences. That's an afternoon's work, and it tackles the biggest gaps first.

And avoid the three mistakes that waste the most time. Don't churn out dozens of thin AI-written posts hoping volume will get you noticed, because models favour sources that say something specific. Don't hide your key facts in PDFs, images or behind forms where they're hard to read. And don't chase one tool's quirks. Build for clarity and consensus, and every answer engine benefits at once.

Do this consistently and you stop hoping the AI mentions you. You engineer it. At LIQID NOISE, this is a core part of our digital marketing work: SEO and AI search visibility built together, so you show up whether your customer types into Google or asks a chatbot.

Frequently asked questions

How do I get my business mentioned by ChatGPT?

Make your business easy to understand and easy to quote. Describe what you do consistently across your website and every profile, publish pages that open with direct answers, earn genuine reviews and mentions on third-party sites, add structured data, and keep key pages current. You can't pay for a mention, but you can give the model clear, well-supported reasons to include you.

Is GEO replacing SEO?

No. GEO builds on SEO rather than replacing it. Many AI tools use search results to find sources, so strong technical SEO, quality content and authority still matter. GEO adds extra priorities on top: quotable, self-contained answers, consistent entity information and a strong presence on third-party sites like review platforms, forums and publications that AI systems draw on.

How long does it take to show up in AI answers?

There's no fixed timeline, because each AI tool updates its sources differently. Some changes, like fixing your pages and profiles, can be picked up by tools that search the live web fairly quickly. Building reviews, mentions and authority takes months of steady work. Treat it like SEO: a compounding investment, measured with a regular monthly prompt check.

Does schema markup help with AI search?

Schema markup helps by turning your key facts into structured, machine-readable data, which makes your business easier to understand and harder to misread. It isn't a magic switch that guarantees citations. It works best alongside clear on-page content that matches the markup, consistent information across the web and genuine third-party mentions.

So what's the difference? SEO gets your pages ranked in search results. AEO gets your content chosen as the direct answer in snippets, AI Overviews and voice assistants. GEO gets your brand named and cited inside the paragraphs AI tools like ChatGPT, Gemini and Perplexity write. Same goal, getting found, across three different kinds of machine.

The good news: you don't need three separate strategies, three agencies and three budgets. You need one set of foundations done properly, then a few specific moves for each. Here's the whole alphabet, explained plainly, with what to actually do about it.

SEO. Search Engine Optimisation

The game you know. You publish pages, earn links, and climb the ranking of ten blue links so a human clicks through. It still matters, but a growing share of searches now end without a single click, because the answer appears before the links do. Similarweb found the share of news searches ending without a click to a news site grew from 56% to nearly 69% in the year after Google's AI Overviews launched.

Don't let the headlines fool you into thinking SEO is dead. It's the foundation the other two stand on. AI answer tools often run searches behind the scenes to find sources, and the pages they find are the pages that are crawlable, fast, well-structured and trusted. If search engines can't find and understand your site, AI tools built on top of search probably can't either.

SEO in 2026 comes down to a few fundamentals:

• Technical health. Pages that load fast, work on mobile, and can be crawled and indexed without errors.
• Search intent. Pages that match what the searcher is actually trying to do, not just the keyword they typed.
• Useful depth. Content that answers the question fully, with real expertise, rather than thin pages written to hit a word count.
• Authority. Links, mentions and reviews from credible sources that tell search engines other people trust you.

AEO. Answer Engine Optimisation

The goal here isn't to rank, it's to be the answer. Featured snippets, Google's AI Overviews, and voice assistants all pull one definitive response to the top and read it aloud or box it off. AEO is engineering your content so yours is the response they lift.

Answer engines are lazy in a very specific way. They want a clean, self-contained answer they can lift without rewriting. So the content that wins AEO tends to look like this:

• Question-shaped headings. Subheadings that match how people actually ask, like "How much does a website cost in Australia?"
• The answer first. One or two plain sentences that directly answer the question, placed immediately under the heading. Detail comes after.
• Lists and steps. Numbered processes and short bulleted lists are easy for machines to extract and easy for humans to scan.
• Structured data. Schema markup that labels your content, such as FAQs, articles, products, services and your organisation, so machines don't have to guess what it is.
• FAQ sections. Real questions from real customers, answered directly.

GEO. Generative Engine Optimisation

This is the newest and fastest-growing front. When someone asks ChatGPT, Gemini, or Perplexity a question, the tool generates a paragraph and names a handful of sources. GEO is the practice of getting your brand woven into that generated answer, cited, recommended, and trusted, before the buyer ever opens a search tab.

GEO is different because the machine isn't picking a page. It's building an opinion. It pulls together what it can find about your brand across your website, reviews, directories, media coverage, forums and other people's articles, then summarises. That means GEO is as much about your reputation across the web as it is about your own content.

The moves that matter most:

• Be a clear entity. Consistent name, description, location and services everywhere your business appears, so AI can tell exactly who you are. More on that in become an entity.
• Write quotable content. Clear definitions, original frameworks, specific numbers from your own work, and strong opinions that are easy to cite. We broke down the style in write content AI can't help but quote.
• Get mentioned elsewhere. Third-party mentions, reviews, podcast appearances and guest articles give AI tools independent evidence to lean on.
• Cover the comparisons. Buyers ask AI "what's the best X for Y." Content that honestly addresses who you're right for, and who you're not, gives the model something to work with.

SEO gets you ranked. AEO gets you answered. GEO gets you generated. The buyer of 2026 touches all three before they touch your website.

Why you can't pick just one

• They share a spine: Clarity, authority, and structured content feed all three at once. Do the foundations right and you compete everywhere.
• The traffic is shifting, not vanishing: Every zero-click answer and AI recommendation is a moment your competitor can own instead of you.
• Early movers get compounding trust: These systems tend to favour brands they already recognise. The presence you build now is the authority they quote for years.

Think about how a real buyer behaves now. They might ask ChatGPT for a shortlist, Google the names it gives them, read an AI Overview on pricing, and then click through to two websites to compare. That's GEO, then SEO, then AEO, then SEO again, in one purchase. Neglect any stage and you drop out of the running without ever knowing you were in it.

One piece of content, built for all three

Here's the practical part. You don't need to write three versions of everything. You need to structure one great page so all three kinds of engine can use it. The checklist:

1. Pick one real question. Something your customers genuinely ask before they buy. Your sales inbox is the best keyword tool you own.
2. Answer it in the first two sentences. Plain, direct, no warm-up. This is your AEO and GEO bait.
3. Then go deep. Explain the how, the cost, the trade-offs and the mistakes. Depth is what earns rankings and trust.
4. Use clear headings. Each subheading should work as a question or a clear label, so machines can pull sections independently.
5. Add proof. Your own experience, specific processes, examples and a named author with real credentials.
6. Mark it up. Add article and FAQ schema, and make sure your organisation details are consistent.
7. Link it up. Connect it to related pages on your site so search engines understand the topic cluster.
8. Keep it current. Update it when things change, and show the date. Stale answers get replaced.

A worked example: the Adelaide physio

Say you run a physiotherapy clinic in Adelaide. Right now your website has a homepage, a services page and a contact form. It ranks for your clinic name and not much else.

You start with one page: "How many physio sessions do you need for lower back pain?" The first two sentences give a straight, honest answer that it depends on the cause and severity, and what a typical course of treatment looks like at your clinic. Below it, you explain the factors, what happens in each session, the warning signs that need a GP, and what patients can do at home. Your principal physio is named as the author, with their qualifications. The page has FAQ schema and links to your back pain service page.

Then you work on the web around you: consistent details on your Google Business Profile and health directories, steady reviews from real patients that mention the problems you treat, and a guest spot on a local health podcast. Over time, that page has a real chance to rank, be lifted into an answer box and get your clinic named when someone asks an AI tool for a good back pain physio nearby. One page, three engines.

Notice what the clinic didn't do. It didn't write fifty thin pages targeting every suburb. It didn't buy links or churn out AI-written posts. It picked the questions patients actually ask, answered them better than anyone nearby, and made sure the rest of the web backed up its claims. Repeat that for the next five questions and you have a content engine that feeds all three letters at once.

How to measure it

This is where most businesses fall down, because the old metrics don't capture it all. Track these:

• Rankings and clicks in Google Search Console, for SEO.
• Impressions without clicks, which can signal you're being seen in answers even when nobody clicks.
• AI referral traffic in your analytics, from tools like ChatGPT and Perplexity.
• Manual AI checks. Once a month, ask the major AI assistants the questions your buyers ask and record whether you're mentioned, and how. We explain why this matters in your customers are asking AI about you.
• "How did you hear about us?" Add AI assistants as an option on your enquiry form. The answers will surprise you.

Mistakes to avoid

• Chasing tricks. Hidden text, keyword stuffing and fake reviews backfire on every engine.
• Treating GEO as separate from SEO. They share foundations. Fix the basics first.
• Burying the answer. Three paragraphs of intro before the point gets you skipped by humans and machines.
• Ignoring your reputation off-site. AI tools read what others say about you, not just what you say.
• Mass-producing AI content. Generic pages add nothing an AI tool can't already say itself, so there's no reason to cite you.

Master all three letters

The alphabet of getting found has three letters now. Master all of them, or watch the answer be about someone else.

This is what our digital marketing team does every day: technical SEO, answer-ready content and the entity and reputation work that gets brands named in AI answers. If you're not sure how you show up across all three, start by asking an AI assistant about your category and seeing whose name comes back. Then let's talk about making it yours.

Frequently asked questions

What is the difference between AEO and GEO?

AEO, answer engine optimisation, is about getting your content chosen as the single direct answer in places like featured snippets, Google's AI Overviews and voice assistants. GEO, generative engine optimisation, is about getting your brand mentioned and cited inside the answers AI tools like ChatGPT, Gemini and Perplexity write. AEO targets extracted answers. GEO targets generated ones.

Is SEO still worth doing now that AI answers questions directly?

Yes. SEO is the foundation for AEO and GEO. AI answer tools often search the web to find sources, and they tend to find pages that are technically sound, well-structured and trusted. Many buyers also still click through to websites before they buy. Dropping SEO would weaken your chances of appearing in AI answers, not improve them.

How do I know if ChatGPT or other AI tools mention my business?

Ask them. Regularly type the questions your customers ask, such as the best provider for your service in your city, into the major AI assistants and record whether you're mentioned and what they say. Also check your website analytics for referral traffic from AI tools, and ask new enquiries where they first heard about you.

How long does AEO and GEO take to work?

There's no fixed timeline. Structural fixes like clear answers, headings and schema can be picked up as soon as pages are recrawled. Building the authority and reputation that makes AI tools trust and recommend you takes longer, usually months of consistent content, reviews and mentions. Treat it as a compounding investment, not a quick switch.

The short answer: use AI for research, rough drafts and variations, then have a human own the strategy, the final voice and every fact. The machine does the typing. You do the thinking. Everything below is how to make that split work day to day, including the exact phrases to hunt down and delete before anything goes live.

Why AI copy sounds like a robot wrote it

It's not a mystery. Language models write by predicting the most likely next word, based on everything they were trained on. The most likely next word is, by definition, the most average one. So left alone, AI drifts toward the safest, most common, most frequently seen phrasing in marketing history.

That's why every unedited AI draft has the same fingerprints:

• Throat-clearing openers that spend a whole paragraph saying nothing before the point arrives.
• Hedged, balanced everything. Every claim softened, every opinion given "on the other hand". No spine.
• Lists of three stacked on lists of three, with a tidy rhythm no real person talks in.
• Inflated words doing the job of plain ones. "Leverage" instead of "use". "Elevate" instead of "improve".
• Zero specifics. No names, no numbers from your business, no stories, no opinions. Nothing only you could have written.

Your customers don't need to know how a language model works to spot this. They've read enough of it now that their brain files it under "ad" and scrolls on.

Where AI genuinely earns its keep

• Research and synthesis: Summarising reviews, mining forums for the exact words your customers use, digesting a competitor's entire funnel in minutes. This is where AI is close to superhuman.
• The ugly first draft: A blank page is expensive. Let AI vomit out a rough version so you have something to react to, cut, and reshape.
• Personalisation at scale: Tailoring the same core message to a dozen segments, work that used to be economically impossible, is now trivial.
• Variant testing: Twenty subject lines, fifteen ad hooks, ten CTAs, generated in seconds so you can test what actually converts.

Notice what all four have in common. In each case, a human still decides what's good. AI widens the options. You pick. That's the pattern to protect.

The research use is the most underrated. Paste in fifty of your own reviews and ask for the phrases customers repeat when they describe the problem you solved. Those phrases are gold. They're the words your next customer is typing into Google and saying to their partner. Copy built from real customer language sounds human because it literally came from humans.

Where a human has to stay in the seat

The failure mode is treating AI as an autopilot instead of a co-pilot. Keep a person firmly in control of the parts that carry your credibility.

Let the machine do the typing. Never let it do the thinking.

• The core strategy and offer: What you promise and to whom is a judgment call, not a generation task.
• The final voice: Rewrite every draft in your own cadence. Cut the throat-clearing. Add the specific, the surprising, the human.
• Every fact and claim: AI hallucinates with total confidence. Ship nothing you haven't verified yourself.

That last one deserves a warning label. AI will invent a statistic, attribute it to a study that doesn't exist, and phrase it so confidently you'll nearly believe it. Publish that under your brand and you've handed a sceptical customer the perfect reason to doubt everything else you say. Regulators don't accept "the AI wrote it" as an excuse for a false claim either. If you can't find the original source yourself, cut it.

Strategy is the other non-negotiable. If you ask AI to decide your positioning, you get the positioning of everyone who asked the same question. We explain why that multiplies whatever you feed it in AI won't replace your marketing, it'll scale what you have.

The workflow: from AI draft to human voice

Here's a repeatable process that keeps the speed and kills the robot. Use it for emails, ads, landing pages and posts.

1. Brief it like a new hire. Give AI your offer, your audience, your one promise, three examples of your best past writing, and the customer phrases you've collected. Garbage brief, garbage draft.
2. Ask for options, not answers. "Give me ten opening lines" beats "write me an email". You want raw material to choose from, not a finished piece to accept.
3. Pick, then rewrite by hand. Choose the strongest bits and rewrite them in your own words. Out loud if you have to. If you wouldn't say it to a customer across a table, it doesn't go in.
4. Add what only you know. A real story, a mistake you made, a specific result, an opinion your competitors wouldn't dare publish. This is the part no model can supply.
5. Run the kill list. Search the draft for the robot phrases below and delete every one.
6. Fact-check line by line. Every number, name and claim gets checked against a source you trust.
7. Read it aloud before it ships. Your ear catches stiffness your eye misses. If you stumble, rewrite it.

The robot-phrase kill list

These phrases are the loudest tells. Find them, cut them, and say the actual thing instead.

• "In a world where..." Start with the point. Nobody needs the world explained to them.
• "Whether you're a... or a..." Pick who you're talking to. Talking to everyone reads as talking to no one.
• "Unleash", "elevate", "empower", "seamless", "cutting-edge". Swap for plain verbs and concrete claims.
• "It's not just X, it's Y." A crutch pattern AI leans on constantly. Once is fine. Three times in a page is a signature.
• "We're passionate about..." Everyone says it. Show what you do differently instead.
• Closing summaries that repeat the whole piece. End on the ask, or on the sharpest line, and stop.

A quick test: if a sentence could appear on any competitor's website without changing a word, it isn't doing any work for you. Rewrite it until only you could have written it. This matters for search too. Generic copy doesn't get quoted, because there's nothing in it worth quoting. The fix is specific, well-structured writing, which we break down in three edits that get your business quoted by ChatGPT.

A worked example: the same email, twice

Here's a hypothetical. Say you run a Perth bookkeeping practice and you want an email inviting past clients to book an end of financial year review.

The unedited AI version opens with "In today's busy world, staying on top of your finances has never been more important." It promises "comprehensive, tailored solutions" and closes with "We look forward to empowering your financial journey." Every word is correct. None of it sounds like you, and none of it gives a busy owner a reason to reply.

Now run the workflow. You feed AI the real phrases clients use, like "I just want someone to tell me what to do before June". You ask for ten subject lines and pick the one that sounds like a person. Then you rewrite the body in two short paragraphs: last year three clients found deductions they'd missed (only if that's true, and checked), you've got eight review slots in the last fortnight of June, reply "yes" and you'll send times.

Same tool. Same fifteen minutes. One reads like a template. The other reads like the person who does their books, because it was.

How to tell if your AI-assisted copy is working

Don't judge it by how polished it looks. Judge it by what people do.

• Replies and responses. For email, the best signal is people writing back. Human-sounding emails get replies. Templates get ignored.
• Click-through against your old copy. Run the AI-assisted version against your best-performing human-written version. Keep whichever wins, not whichever was faster to make.
• Unsubscribes and hides. A rise after you switched to AI drafting is a warning that the voice has gone generic.
• What customers say back to you. If prospects start quoting your lines on sales calls, the copy is landing. If nobody remembers anything you wrote, it isn't.

Give each test enough volume and time to be readable, change one thing at a time, and write down what you learned. Over a few months you'll build a private playbook of what your audience responds to, which becomes the best brief you can ever give an AI tool.

Voice is a strategy decision, not a style setting

Here's the deeper point. You can't edit your way to a distinctive voice if you don't know what you stand for. The robotic sound is often a symptom, not the disease. A business that hasn't decided who it's for and what it believes will sound generic whether a human or a machine writes the copy.

So before you worry about prompts, get clear on three things: who you're talking to, what you believe that your competitors don't, and how you'd explain your offer to a friend at a barbecue. That barbecue version is your voice. Write it down, keep examples of it, and feed them to AI every time. If you're struggling to pin it down, your message might be the real problem, not your tools.

The 10x rule

Used well, AI doesn't replace your marketer, it turns one sharp marketer into the output of a small team. The machine handles the volume, the human guards the taste. That combination moves far faster than either could alone, and it still sounds like a person wrote it. Because one did.

That's how we use it at LIQID NOISE across our digital marketing work: AI for speed and range, humans for strategy, voice and truth. If you want marketing that moves at machine speed and still sounds unmistakably like you, that's the balance to build.

Frequently asked questions

How do I make ChatGPT writing sound more human?

Give it better inputs and edit harder. Feed it real customer phrases and examples of your own writing, ask for several options instead of one finished draft, then rewrite the best bits in your own words. Cut inflated words and generic openers, add a real story or opinion, and read it aloud before publishing.

Can customers tell when marketing is written by AI?

Often, yes. Not because they detect the tool, but because unedited AI copy is generic, over-polished and free of specifics. People have read so much of it that the pattern registers as "ad" and they move on. Copy with real details, opinions and customer language reads as human regardless of how the first draft was made.

What words make writing sound like AI?

The biggest tells are inflated verbs like "unleash", "elevate" and "empower", filler openers like "in a world where", the "it's not just X, it's Y" pattern used over and over, and tidy lists of three everywhere. None are banned on their own. The problem is density. Swap them for plain words and concrete claims and the copy instantly sounds more human.

Is it OK to use AI to write marketing copy for my business?

Yes, as long as a human owns the result. Use AI for research, drafts and variations to save time. Keep a person responsible for the strategy, the final wording and checking every fact and claim. You're accountable for what you publish, whoever or whatever typed the first version.

That is exactly what most websites do. They look pretty. They "represent the brand." And they quietly cost you customers every single day while you pat yourself on the back for having a nice logo up top.

Here's the fix in plain terms: a website that sells leads with the visitor's problem, makes one next step obvious, puts proof exactly where doubt shows up, removes every bit of friction, and never lets an interested visitor leave without a way to raise their hand. Do those five things and the same traffic you already have starts turning into enquiries.

Pretty is not the job. Closing is the job.

Your website has one job: take a stranger who has never heard of you and walk them, step by step, to a decision. Everything else, the animations, the clever tagline, the stock photo of people high-fiving, is set dressing. If it does not move the visitor closer to buying, it is working against you.

A website that does not sell is not a brand asset. It is a very expensive business card that nobody keeps.

The brutal truth is that visitors are not reading. They are scanning, deciding in seconds whether you are worth their time, and leaving the moment they get confused. Confusion is the enemy. Confused people never buy.

Think about what a great salesperson does in the first minute with a new prospect. They find out what the person needs. They show they understand the problem. They explain, simply, how they can help. They share a story about someone in the same situation. Then they suggest a next step. Your website should do exactly the same thing, in the same order, without anyone in the room.

Why most websites fail at selling

Most websites are briefed like brochures. The conversation with the designer is about colours, fonts, the logo and which photos to use. Nobody asks "what does a visitor need to believe before they'll get in touch, and where on the page do we prove it?"

So you end up with a site organised around the business, not the buyer. A homepage that opens with "Welcome to Smith & Co, established 1998." An About page that's the longest page on the site. Services buried three clicks deep. A contact form at the very bottom of one page, asking for eleven fields.

It's not that anyone made a bad decision. It's that nobody made the sales decision at all. And cheap builds make it worse, because there's no budget for strategy, only for templates. We break down that hidden cost in the real cost of a cheap website.

What a website that actually sells does differently

• It leads with the visitor, not with you. The headline is about their problem and the outcome they want, not your company history.
• It makes one thing obvious: what to do next. One primary action, repeated, impossible to miss.
• It stacks proof, results, testimonials, logos, numbers, exactly where doubt creeps in.
• It removes friction at every step. Fewer fields, fewer clicks, fewer reasons to bounce.
• It never lets a warm visitor leave empty-handed without an invitation to raise their hand.

Let's make each one concrete.

Leading with the visitor means your headline passes the "so what?" test. "Award-winning landscape design" fails. "Turn your backyard into the room you actually use" passes. The subheading then says who it's for and how it works, in one line.

One obvious next step means choosing a single primary action, like "Book a free quote," and repeating it in the header, after every major section and at the bottom. Secondary options are fine, but they should look secondary.

Stacking proof means placing a testimonial right beside the claim it supports, not hiding them all on a separate page. If visitors worry about price, put a review about value next to your pricing. If they worry about timelines, put a review about finishing on time next to your process.

Removing friction means asking only for what you need to start the conversation. Name, email or phone, and a short message is usually enough. It also means speed. Every second of loading is a small reason to leave. We cover that in detail in a slow website is a silent business killer.

Catching warm visitors means having something for people who aren't ready to call yet. A price guide, a checklist, a short quiz or a newsletter. Most visitors won't buy on their first visit. The ones who leave an email can be followed up.

Do this and your traffic does not need to change at all. The same number of people arrive, but instead of leaking away, they convert. That is the difference between a website that costs you money and one that makes it.

The sales structure of a homepage, top to bottom

You don't need to reinvent the wheel. A homepage that sells usually follows the same path a good sales conversation does:

1. The hook. A headline about the visitor's outcome and a subheading about who it's for. One primary button.
2. The problem. A short section that names the frustration they're living with, in their words.
3. The solution. What you do, broken into clear services or steps. Short, scannable, linked to deeper pages.
4. The proof. Reviews, results, before-and-afters, recognisable client logos or credentials, whatever is most convincing in your industry.
5. The process. Three or four steps showing what happens after they get in touch. This lowers the fear of the unknown.
6. The objections. A short FAQ that answers the questions that stop people buying, like price, timeframes and what's included.
7. The close. Repeat the offer and the primary action. Make it easy.

Service pages and campaign pages follow the same logic, just more focused. For the page you send paid traffic to, read the anatomy of a landing page that prints money.

A worked example: the landscaping business

Say you run a landscaping business on the Gold Coast. Your website gets a steady flow of visitors from Google and Instagram. Enquiries are slow.

Right now, the homepage opens with a full-screen slideshow and the words "Welcome to Coastal Landscapes." There's a paragraph about your history, a grid of project photos with no descriptions, and a contact form at the bottom of the Contact page that asks for budget, address, project type, timeframe and how you heard about the business.

You rebuild it as a salesperson. The headline becomes "Backyards built for the way you actually live," with a subheading naming your service area and a "Get a free design consult" button. Below that, a short section names the problem: a yard you never use, quotes that blow out, tradies who vanish mid-job. Then your three core services, each with a before-and-after and a one-line review. A four-step process. An FAQ about cost ranges and timeframes. The form shrinks to name, phone, suburb and a message, and it appears on every page.

Nothing about your traffic changed. But visitors now understand what you do in seconds, see proof right where they hesitate, and have an easy way to take the next step. That's the shift from brochure to salesperson. (The business is hypothetical, but these are exactly the changes that make the difference.)

How to measure whether your website is selling

You can't improve what you don't track. At a minimum, set up:

• Conversion tracking on every form submission, phone number click and booking, so you know how many visitors take action.
• Conversion rate by page, so you can see which pages persuade and which pages leak.
• Lead source, so you know whether enquiries came from search, social, ads or referrals.
• Lead quality, logged by whoever answers the enquiries. A site that brings more of the wrong leads isn't a win.

Then watch recordings or heatmaps for a week if your analytics tool offers them. You'll see exactly where people stall, what they ignore and where they give up. It's one of the fastest ways to find your biggest leak.

One more place to check: what AI tools say about you. More buyers now ask a chatbot before they visit a website at all, and the answer often draws on what your site says. Clear pages with direct answers help there too. See your customers are asking AI about you.

Mistakes that keep your salesperson asleep

• Redesigning without rewriting. A new look with the same vague copy is the same salesperson in a nicer suit.
• Too many calls to action. "Call us, email us, follow us, download this, subscribe" on one screen means the visitor does none of them.
• Hiding prices entirely. You may not be able to list exact prices, but a starting point or typical range answers the question every buyer is asking and filters out poor fits.
• Ignoring mobile. Plenty of people will meet your site on a phone, around 43% of Australian web traffic in August 2026 according to StatCounter. Check every page, every button and every form on one before you call it finished.
• Launching and forgetting. A salesperson gets coached. So should your website. Review the numbers monthly and keep improving.

Your competitors are counting on you doing nothing

Every day you run a site that informs but does not persuade, the business down the street with a sharper, faster, more convincing website is eating clicks that should have been yours. Search does not reward the oldest company. It rewards the one that turns attention into action.

The good news is that most websites in most industries are still brochures. The bar is low. A site that simply answers the buyer's questions, shows proof and asks for the next step clearly will stand out in almost any local market.

We build sites engineered around one metric that matters: did the visitor take the next step? Everything we design, write, and wire up bends toward that answer. That's the core of our web development work, strategy and copy first, design and code in service of them. If you are ready to stop running a brochure and start running a salesperson that never sleeps, never calls in sick, and never lets a buyer slip away, let's talk.

Frequently asked questions

Why is my website getting traffic but no enquiries?

Traffic without enquiries usually means visitors can't quickly tell what you do, don't see enough proof, can't find an obvious next step, or hit friction like a slow page or a long form. Check your headline, make one primary action visible throughout the page, place reviews next to key claims and shorten your forms. Then track conversions by page to find the biggest leak.

What should be above the fold on a homepage?

Above the fold should answer three questions in seconds: what you do, who it's for and what to do next. That means an outcome-focused headline, a one-line subheading and a clear primary button, ideally with a small piece of proof like a rating or short review. Save your company history and detailed service lists for further down the page.

How many fields should a contact form have?

As few as you need to start the conversation. For most service businesses, name, email or phone, and a short message is enough. Every extra field is another reason to give up. If you need more detail to qualify leads, gather it on the first call or in a follow-up, rather than making strangers do the work before they trust you.

What is a good website conversion rate?

It depends heavily on your industry, price point and traffic source, so broad benchmarks can mislead. The most useful comparison is your own site over time. Track conversion rate by page and by source, make one meaningful change at a time and measure the difference. A steady improvement on your own numbers matters far more than matching someone else's average.

So how do you stand out when AI can write anything? You stop competing on quantity and start competing on the things a machine can't produce: a real point of view, experience only you have, and a voice people recognise without seeing your logo. Distinctiveness is the only life raft. Everything else sinks with the flood.

Here's why the old playbook just stopped working, and exactly what to do instead.

When everything is easy, nothing is impressive

The moment infinite content became free to produce, its value collapsed. Ten thousand "5 Tips to Grow Your Business" posts don't build a brand, they build background noise. And the algorithms, trained on that same slop, are getting very good at recognising and burying it.

Think about what that does to the standard content strategy. For a decade the advice was simple: publish often, cover every keyword, be everywhere. That worked when publishing was hard, because effort itself was a filter. The businesses willing to grind out two posts a week beat the ones that couldn't be bothered.

That filter is gone. Your competitor can now generate a month of posts before lunch. So can their competitor. So can the kid in his bedroom who's never run a business. When everyone can produce the same volume, volume stops meaning anything, to readers, to search engines and to the AI assistants that now summarise the web for your customers.

You cannot out-publish a machine. You can only out-mean one.

How audiences learned to skip the slop

Readers adapt fast. Plenty of people now recognise the tells of lazy AI writing within a line or two. The breathless opener. The three tidy bullet points with bolded labels that say nothing. The neat summary at the end that repeats the intro. The total absence of an opinion anyone could disagree with.

Once a reader spots those tells, they stop reading and start skimming, and they quietly downgrade their opinion of whoever published it. The cost isn't just a wasted post. It's the slow erosion of trust in your brand every time someone scrolls past something that sounded like everybody else.

And here's the twist. The same thing is happening one layer up. AI answer engines pulling information to answer your customers' questions have no reason to quote the thousandth generic take on a topic. They look for sources that add something. We've broken down what makes content AI can't help but quote, and "more of the same" isn't on the list.

Your point of view is the one thing that can't be prompted

A large language model can imitate any style, but it cannot manufacture conviction. It has no scars, no strong opinions, no stake in the outcome. That is precisely where your moat lives.

• A real POV: Take a position your competitors are too scared to take. Generic advice is now infinite and worthless, a sharp opinion is rare and magnetic.
• Proprietary experience: Your data, your client results, your war stories, the stuff that only exists because you lived it, can't be scraped or synthesised.
• A distinctive voice: When AI flattens everyone toward the same competent, forgettable middle, sounding unmistakably like you becomes a superpower.

Each of these works for the same reason. A model learns from what already exists and predicts the most likely next word. By design, it pulls toward the average. Anything that sits away from the average, a contrarian call, a story from last Tuesday, a phrase only you would use, is something it can't produce on its own. That gap is your advantage, and it gets more valuable as the flood gets deeper.

How to find your distinctive angle

Most businesses say they have a point of view. Few can state it in a sentence. Here's a practical way to dig yours out:

1. List the advice in your industry that you think is wrong. Every field has sacred cows. Where do you quietly disagree with the consensus? That disagreement is content nobody else can write.
2. Write down the questions clients ask you most. Then write down the answer you actually give in the room, not the polite version on your website. The real answer is usually the interesting one.
3. Mine your mistakes. What did you get wrong early on? What did it cost? Honest failure stories are the most human content there is, and the most trusted.
4. Collect your own numbers. Response times, project timelines, common costs, patterns you see across jobs. Even simple observations from your own work are information no model has.
5. Name your method. If you have a way of doing things, give it steps and a name. A named framework is memorable, repeatable and yours.

Do this exercise once and you'll have more material than you can publish in a year. None of it can be copied by a competitor with a prompt, because none of it existed until you wrote it down.

A worked example: two plumbers, one topic

Here's a hypothetical. Two plumbing businesses in the same city both want to rank for "hot water system replacement cost."

The first asks an AI tool for a blog post. It gets a perfectly competent article: factors that affect cost, types of systems, a reminder to get multiple quotes. It reads exactly like the forty other articles on the same page of results, because it was built from them.

The second business writes from the jobs it's actually done. It explains why the cheapest quote often leads to a callback in a few years, walks through the three situations where they'd tell a customer to repair instead of replace, and admits which system they'd put in their own house and why. It includes the question customers forget to ask until it's too late.

Which one would you trust? Which one would an AI assistant quote when someone asks what to watch out for? Which one makes the phone ring? The second post took longer to write. It's also the only one worth publishing.

Use AI to sharpen, not to fill

The answer isn't to abandon AI, it's to stop using it as a content vending machine. Use it to research faster, to pressure-test your arguments, to draft so you can spend your energy editing for edge. Let the machine handle the plumbing. You handle the point.

In practice, a sharpening workflow looks like this:

• Start with your raw thinking. Voice notes, bullet points, a rant after a client call. The substance comes from you first.
• Use AI to structure it. Ask it to organise your notes into an outline, or to spot gaps in the argument.
• Ask it to argue back. Have it play a sceptical customer and poke holes in your position. Then strengthen the weak points.
• Edit ruthlessly by hand. Cut every generic sentence. Replace every vague claim with a specific example. Make it sound like you talking.
• Let it handle the grunt work. Repurposing into captions, summarising for email, checking for typos. That's where speed actually pays.

We've written a full guide on using AI in your marketing without sounding like a robot. The short version: the machine is a fantastic assistant and a terrible author.

Common mistakes that drown brands in the flood

• Publishing to a quota. "Three posts a week" is a production target, not a strategy. One post worth reading beats ten nobody finishes.
• Hedging every opinion. Content that tries to offend no one persuades no one. Take a side.
• Hiding the humans. Faceless brand content is exactly what AI produces best. Put real people, names and stories front and centre.
• Copying the top result. If your plan is to write what's already ranking, only slightly better, a machine can already do that.
• Measuring output instead of outcomes. Count enquiries, replies, shares and mentions, not the number of posts you shipped.

What to do this week

Pull up the last ten pieces of content you published. Read each one and ask a brutal question: could a competitor have published this word for word with their logo on it? If the answer is yes, it isn't building your brand. It's filling space.

Then pick the one topic your customers ask about most and rewrite that piece from scratch using the exercise above. Add one opinion, one real story, one specific number from your own work and one thing you'd tell a friend that you wouldn't normally say in public. That single piece will do more for you than a month of volume.

Finally, change how you keep score. Stop reporting on how many posts went out. Start tracking the signals that tell you content is actually landing: replies to your emails, enquiries that mention a specific article, time spent on the page, people forwarding your work to a colleague, and whether your brand name gets searched more often over the following months. Those numbers move slowly, but they move for distinctive content and they barely twitch for filler. Give the new approach a full quarter before you judge it.

Distinctiveness is the whole game

The flood isn't going to recede. It's going to get deeper every month as the tools get cheaper and faster. That's bad news for anyone whose marketing plan was "publish more." It's great news for anyone willing to be specific, opinionated and human, because the contrast has never been sharper.

When every brand can generate competent copy, competent copy is worth nothing. What's left to buy is who you are, and we've argued that brand is the last moat for exactly this reason.

In a world drowning in average, being unmistakably, memorably yourself isn't a nice-to-have. It's the whole game. If you want help finding your angle and building a content engine around it, our digital marketing team does exactly that.

Frequently asked questions

Does Google penalise AI-generated content?

Google's own guidance says it rewards helpful, original, high-quality content however it is produced. The real risk with AI content is that it tends to be generic, and generic pages struggle to rank or get cited because they add nothing new. Content that uses AI for efficiency but carries real expertise, experience and opinion is a different thing entirely.

How do I make my content stand out from AI content?

Add what a model can't: your opinions, your own stories and data, named people, specific examples from real work and a recognisable voice. Take positions your competitors avoid. Answer the questions customers actually ask you, the way you'd answer them in person. If a competitor could publish your piece unchanged, it isn't distinctive enough yet.

Should small businesses still use AI for marketing?

Yes, but as an assistant rather than an author. Use it for research, outlining, pressure-testing arguments, repurposing content into different formats and cleaning up drafts. Keep the thinking, opinions and final edit human. That combination gives you the speed of AI without the sameness that makes AI-only content easy to ignore.

How often should I publish content now?

As often as you can publish something genuinely worth reading, and no more. A steady rhythm matters for building an audience, but quality now beats frequency. One distinctive, useful piece a fortnight will usually outperform daily generic posts, both with readers and with the search and AI systems deciding what to show them.

Nobody moves in a neat line anymore. They see you on a reel, forget you, get retargeted, ask a friend, read a review, disappear for a month, then buy at 11pm on their phone because something reminded them you exist. The funnel is not a funnel. It is a messy loop, and if your marketing still assumes a straight line, you are losing people in the gaps.

So what is replacing the funnel? A flywheel: a connected system where attention builds trust, trust drives sales, and happy customers generate fresh attention. Instead of pushing strangers down a chute one campaign at a time, you build something that keeps turning and gets cheaper to run the longer it spins.

Why the funnel stopped matching reality

The funnel was drawn for a world with a few channels and a predictable path. You saw an ad, you visited a store, you bought. Marketers could reasonably assume each stage led to the next.

That world is gone. Your buyer now bounces between short video, search, AI assistants, group chats, review sites, email and your website, in no particular order and often across weeks. They go backwards. They skip stages. They are "aware" of you three separate times before anything sticks. They get to the brink of buying, go quiet, and come back after a friend mentions you in passing.

A funnel treats all of that as leakage. It measures how many people fell out of each stage and tries to stop the leaks with more spend at the top. The result is a business that is permanently buying new attention to replace attention it failed to keep.

From funnel to flywheel

The brands winning now do not push people down a chute. They build a flywheel: a system where attention feeds trust, trust feeds sales, and happy customers feed more attention. It spins faster the longer it runs, and it compounds instead of resetting to zero every campaign.

The difference is not a new diagram. It is a different question. A funnel asks "how do we push more people through?" A flywheel asks "what makes this easier to turn every month?" That question changes where you spend. Suddenly the customer experience, the follow-up, the review request and the referral program matter just as much as the ad budget, because they are what keep the wheel moving without you pushing it.

Stop thinking in campaigns with a start and an end. Start building a machine that never stops turning.

What the loop actually needs

• Constant presence so you are already familiar when the buying moment hits, not scrambling to introduce yourself.
• Trust assets, content, proof, and reputation working around the clock whether or not you are running ads.
• Frictionless capture so the moment someone raises a hand, you catch it.
• Follow-up that does not quit, because plenty of buyers take longer to decide than most marketers are willing to wait.
• A reason for happy customers to talk, feeding fresh attention back into the top.

Here is what each part looks like when it is built properly.

Constant presence

You do not need to be everywhere. You need to be consistently in the two or three places your buyers actually spend time, showing up often enough that you are the obvious name when the need arrives. A steady rhythm of useful content beats a burst of activity followed by silence. We cover how to set that rhythm up in build a content engine, not a content calendar.

Trust assets

Reviews, case studies, detailed answers to common questions, videos of real customers, a founder who shows their face. These work every hour of every day, including the 11pm moment when a buyer is deciding. They also get found by search engines and AI assistants long after they are published.

Frictionless capture

When someone is ready to act, the path must be obvious and short. A clear form, a booking link, an email signup with a genuine reason to join, a phone number that gets answered. Every extra step is a gap in the loop.

Follow-up that does not quit

Most buyers are not ready the first time they raise their hand. The flywheel keeps them warm with email, helpful content and respectful retargeting, so when they are ready, you are the one they come back to. Email in particular is the cheapest, most owned follow-up channel you have, which is why we call it the highest-ROI channel nobody respects.

A reason to talk

Happy customers are the flywheel's engine. Make it easy and natural for them to review you, refer friends and share what they bought. Ask at the right moment, when the result is fresh, not weeks later.

Why this crushes cost-per-sale

When every part reinforces the others, you stop paying full price for every single customer. Your content warms up your ads. Your ads feed your list. Your list creates referrals. Each customer costs less than the last because the machine is doing work your budget used to do alone.

Think of it as friction and momentum. In a funnel, every new customer needs a fresh push from the ad budget. In a flywheel, some customers arrive already warm because they have seen your content, been referred or read your reviews. They tend to convert faster and need fewer ad impressions. As that share grows, your average cost per sale falls, even if the cost of cold traffic stays exactly the same.

How to build your flywheel, step by step

1. Map the real journey. Ask your last ten customers how they found you, what they checked before buying, and what nearly stopped them. You will find loops and detours the funnel diagram never showed.
2. Find the biggest gap. Where do people drop out and never come back? No follow-up after an enquiry? No way to capture interested visitors who are not ready yet? Fix the leakiest point first.
3. Own the audience. Build an email list and customer database you control, so you are not renting every touch from an ad platform. Our case for this is in rent your audience and you will lose it.
4. Set up automatic follow-up. A welcome sequence for new subscribers, a nurture sequence for enquiries that went cold, and a check-in after purchase.
5. Systemise reviews and referrals. A request sent at the moment of delight, every time, without anyone having to remember.
6. Feed it with content and ads. Only now does spending on attention pay properly, because the rest of the machine is ready to catch and keep what you buy.
7. Review monthly. Look at the whole loop, not just the ads. Where is it slowing? That is next month's work.

A worked example: the Sydney landscaper

Say you run a hypothetical landscaping business in Sydney, spending $3,000 a month on Meta ads. The funnel version: ads drive enquiries, you quote, some say yes, the rest vanish. When you pause ads, enquiries dry up within weeks. Every month starts from zero.

Now build the loop. Every finished job gets photographed and turned into a before and after post and a short reel. Every enquiry that does not book gets a simple email sequence with project examples, seasonal tips and a no-pressure invitation to reconnect. Every happy client gets a review request the day the job is finished and a thank-you gift card for any neighbour they refer. Your ads now retarget people who watched your reels, rather than only strangers.

A few months in, some enquiries mention a neighbour, some come from people who got your emails for weeks before booking, and some found your reviews first. The ad spend has not changed. What changed is that the ads are no longer doing all the work alone. That is the flywheel.

How to measure a flywheel instead of a funnel

Funnel metrics look at stages in isolation. Flywheel metrics look at momentum.

• Blended cost per customer. Total marketing spend divided by all new customers, not just ones an ad platform claims credit for.
• Share of warm customers. The proportion who arrived through referral, repeat purchase, email or direct visits. Rising is good.
• Time to purchase. Are buyers deciding faster because they already trust you?
• Review velocity. How many new reviews each month. It is a direct read on whether happy customers are talking.
• Repeat and referral revenue. The clearest sign the wheel is turning on its own.

A useful gut check: what happens to revenue if you pause ads for a month? In a funnel business, it falls off a cliff. In a flywheel business, it dips, and the rest of the system keeps turning.

Common mistakes when ditching the funnel

• Renaming, not rebuilding. Calling your funnel a flywheel on a slide changes nothing.
• Neglecting existing customers. The wheel is powered by people who already bought. Ignore them and it stalls.
• Stalking instead of following up. Relentless retargeting burns goodwill. We show a better way in retargeting that persuades without stalking.
• Trusting last-click credit. In a loop, the last touch rarely did all the work. Measure the whole system.

Stop renting attention, start owning a system

That is the shift: from renting attention one campaign at a time to owning a system that compounds. We build these loops for ambitious brands, mapping every touch, sealing every gap, and turning scattered marketing into one machine that spins faster every month. Our digital marketing team starts with the journey your buyers actually take, not the one on the slide. If your marketing resets to zero every time you stop spending, you do not have a system. Let's build you one.

Frequently asked questions

Is the marketing funnel still relevant in 2026?

As a rough way to think about buyer readiness, it can still be useful. As a model of how people actually buy, it falls short. Real buyers loop, stall, return and get influenced by reviews, referrals and AI assistants along the way. Treating marketing as a connected loop rather than a straight line helps you catch and keep the people a funnel would call leakage.

What is the difference between a funnel and a flywheel?

A funnel treats customers as the output at the bottom of a one-way process. A flywheel treats them as the energy that drives the next cycle. In a flywheel, customers generate reviews, referrals and content that attract new buyers, so momentum builds over time. The funnel needs constant pushing. The flywheel gets easier to turn as it grows.

How do I start building a marketing flywheel for a small business?

Start by asking recent customers how they really found and chose you. Then fix the biggest gap, usually missing follow-up or no way to capture interested people who are not ready yet. Build an email list, automate a simple nurture sequence and ask every happy customer for a review. Add ads once the system can catch and keep what you buy.

How long does a flywheel take to show results?

Some parts pay off quickly. Follow-up emails and review requests can lift results within weeks because they catch people you were already losing. The compounding effect of content, referrals and reputation builds over months. The key is consistency. A flywheel gains momentum slowly at first and then keeps turning with less effort, which is the opposite of campaign spend.

So will AI replace your marketing? No. It will scale whatever marketing you already have. If your offer is strong and your message is sharp, AI lets you produce and test at a volume that used to need a whole department. If your offer is weak and your message is fuzzy, AI just helps you broadcast that weakness to more people, faster, for less money.

That's the uncomfortable truth underneath every shiny tool demo. Let's break down what it means for how you should actually spend your time and budget.

AI is a multiplier, not a magic wand

Point a multiplier at a strong offer, a sharp message, and a hungry audience, and it prints results at a scale you could never reach by hand. Point that same multiplier at a vague value proposition and a "we help businesses grow" tagline, and it will manufacture confusion faster than any junior hire ever could.

AI makes good marketing cheaper and bad marketing louder. The gap between the two is about to become brutally obvious.

Think about the maths of a multiplier. Ten times zero is still zero. Ten times a negative number is a bigger negative number. If your ads currently lose money, generating ten times more ad variations doesn't fix the loss. It just gives the platform more ways to spend your budget on the same broken idea.

And here's the part nobody in the tool demos mentions: your competitors have the same tools. The same models, the same prompts, often the same templates. When execution is available to everyone at almost no cost, execution stops being an advantage. The only thing left to compete on is what goes into the machine.

Why AI exposes weak strategy so fast

Before AI, a vague strategy could hide for years. It took so long and cost so much to produce campaigns that nobody ever ran enough of them to see the pattern. A flat quarter got blamed on the season, the agency, the algorithm.

Now you can launch fifty ad variations in an afternoon. When all fifty flop, there's nowhere left to hide. It wasn't the headline. It wasn't the image. It was the thing underneath all fifty of them, the offer or the message or the market, and that was never the AI's job.

That's actually good news if you're willing to hear it. AI is the fastest diagnostic tool marketing has ever had. It compresses years of slow feedback into weeks. The businesses that treat that feedback honestly get sharper every month. The ones that keep blaming the tool keep generating more of the same.

What actually moves the needle

The fundamentals haven't changed. They've just become the only thing that still matters, because everything downstream of them is now nearly free to produce.

• The offer: No prompt on earth fixes a product nobody wants at a price nobody will pay. Make the offer irresistible first, then let AI pour fuel on it.
• The message: AI can write a thousand headlines, but it cannot decide what your one big promise is. That is a human, strategic call, and it is worth more than ever.
• The market: Speaking to everyone means speaking to no one, and AI will happily help you do exactly that at industrial scale.

Get these three right and almost any competent execution works. Get them wrong and no execution saves you. If your offer is the weak link, start with building an offer people can't refuse before you touch a single AI tool.

The strategy check to run before you scale anything

Before you plug AI into your marketing, run this check. It takes an afternoon and it tells you whether you're about to scale a winner or a leak.

1. Write your offer in one sentence. Who it's for, what they get, and why it beats the alternative. If it takes a paragraph, it isn't ready to scale.
2. Name the one promise. The single biggest outcome you deliver. Not three. One. Every headline AI writes for you should be a variation on this, not a new idea.
3. Describe your best customer in detail. Their job, their frustration, the words they use when they complain about the problem. Pull real phrases from reviews, emails and sales calls.
4. Find proof it already works. Something has to be converting today, even at small volume. A referral source, one ad, one email. Scale what already works. Never scale a guess.
5. Pick one number that matters. Booked calls, sales, qualified leads. Not impressions, not likes, not "engagement". If you can't name it, AI will optimise toward whatever is easiest to count.

If you can answer all five clearly, you're ready to scale. If you stall on any of them, that's where your attention goes first. Not the tools.

A worked example: same tools, opposite results

Let's run a hypothetical. Two Melbourne landscaping businesses each start using AI to write ads, emails and social posts. Same tools, same budget, same month.

Business A sells "quality landscaping services for homes and businesses". It asks AI for thirty Facebook ads. It gets thirty polished, friendly, forgettable ads about quality and service. They look exactly like every other landscaper's ads because they're built from the same generic input. Clicks are cheap, enquiries are few, and the ones that come in are mostly price shoppers.

Business B has done the hard thinking first. It sells one thing: low-maintenance backyard makeovers for busy families, finished in two weeks, with a fixed quote and a promise that they'll be back to fix anything that fails in the first year. It asks AI for thirty ads, every one a different angle on that same promise. Some lead with the fixed price. Some lead with time back on weekends. Some lead with the guarantee. Within a few weeks it knows which angle wins, and it scales that one.

Same technology. Business A scaled a vague idea and got vague results faster. Business B scaled a sharp idea and found its winner faster. The tool didn't decide the outcome. The strategy did.

Notice what Business B actually used AI for. Not to invent the offer. Not to choose the audience. It used AI to explore angles on a decision it had already made, and to find the winner faster than a human copywriter could test by hand. That's the correct division of labour, and it's available to any business willing to do the thinking first.

The teams that win

The winners aren't the ones with the fanciest tools. They're the ones who use AI to compress the boring 80%, the research, the first drafts, the variant testing, the reporting, so their humans are freed up for the 20% that machines can't touch: taste, judgment, positioning, and the nerve to make a bold promise and stand behind it.

In practice, that split looks like this:

• Hand to AI: summarising customer reviews, drafting variations of a proven headline, resizing and reformatting assets, first drafts of emails, turning a sales call transcript into notes, pulling weekly numbers into a report.
• Keep with humans: deciding who you serve and who you don't, setting the price, writing the core promise, choosing which test results to trust, and every claim that goes out with your name on it.

We go deeper on drawing that line in what to automate and what never to. The short version: give AI the work, keep the judgement.

The mistakes that turn AI into a megaphone for mediocrity

Generating before deciding. Opening a chat window and asking for "a marketing plan for my business" gets you the average of every marketing plan ever written. Decide first, then generate.

Publishing volume for its own sake. Ten blog posts a week that say nothing new don't build authority. They bury the one good idea you have under a pile of filler. Audiences and search engines are both getting better at ignoring the flood, as we cover in the AI content flood.

Letting the machine pick the message. If AI writes your positioning, you sound like everyone else who asked the same question. Your message should come from your customers' words and your own conviction.

Measuring output instead of outcomes. "We published 60 posts this month" is not a result. Sales, booked calls and qualified leads are results. Track those, and cut whatever doesn't feed them.

How to tell if AI is actually helping

Run a simple before-and-after. Pick one channel, note your cost per qualified lead or cost per sale for the last three months, then bring AI into the workflow for that channel only. After a couple of months, compare.

Look for two things. First, is the cost per result falling? Second, is your team spending the hours it saved on higher-value work, like talking to customers or sharpening the offer? If output went up but results stayed flat, AI is scaling the wrong thing. Go back to the strategy check above.

Watch quality signals too, not just cost. Are the leads coming in better or worse fit than before? Are customers replying to your emails, or quietly unsubscribing? Is your sales team closing at the same rate? AI can drive a cheaper cost per lead by finding people who are easy to reach and hard to sell. A lower cost per lead that your sales team hates is not a win. It's a bill you pay later.

Aim first, then scale

AI won't take your job. A marketer who knows how to aim it, however, absolutely will. Learn to aim it, or hire someone who already can.

That's the whole game from here. The tools will keep getting cheaper and better for everyone, which means the advantage keeps shifting to whoever has the clearest offer, the sharpest message and the most honest read on their market. That's strategy work, and it's the part of digital marketing we obsess over at LIQID NOISE before we ever switch a machine on. Get the aim right and AI becomes the best hire you'll ever make. Get it wrong and it becomes the most efficient way to waste money ever invented.

Frequently asked questions

Will AI replace marketing agencies and marketing jobs?

AI replaces tasks, not strategy. Drafting, resizing, reporting and research are getting automated fast. Deciding who to sell to, what to promise and what to charge still needs human judgement. The marketers at risk are the ones whose whole job was producing volume. The ones who can aim AI at a sharp strategy become more valuable, not less.

What should I use AI for first in my marketing?

Start with work that is repetitive and low risk: summarising customer reviews for language you can reuse, drafting variations of a headline that already works, and turning raw numbers into a weekly report. These save hours without putting your reputation on the line. Keep your offer, positioning and final claims in human hands.

Is AI-generated content bad for SEO?

Not because a machine helped write it. Search engines reward content that is useful, accurate and genuinely adds something. The trouble is that most unedited AI content is none of those things. It repeats what already ranks and says it blandly. Use AI to speed up drafting, then add your own expertise, examples and opinions before anything goes live.

Why is my AI-generated marketing not converting?

Usually because the problem sits upstream of the content. If the offer is unclear, the promise is generic or the audience is too broad, AI just produces more of that faster. Test the fundamentals first: can you state your offer in one sentence, and is anything converting today? Fix the input before judging the tool.

It is three moves: Hook. Story. Offer. Miss any one and the ad dies. Nail all three and you have something that sells while you sleep.

Put simply, the hook stops the scroll, the story makes the prospect believe you understand their problem and can fix it, and the offer gives them a clear, low-risk reason to act now. It works on a fifteen-second vertical video, a long-form sales page, a radio spot or a cold email, because it follows the order people actually move through when they decide to buy. Here's how to build each part, step by step, and how to tell which one is broken when an ad flops.

The Hook: earn the next three seconds

Nobody owes your ad attention. In the scroll, you have a heartbeat to stop the thumb. The hook is a pattern interrupt, a bold claim, a sharp question, a visual that does not belong. Its only job is to buy you the next three seconds. Weak hook, and it does not matter how brilliant the rest is. Nobody saw it.

You are not competing with other ads. You are competing with your prospect's best friend, their kid, and a cat video. Hook like it.

Most hooks fail for one reason: they're about the business, not the viewer. "Introducing our new spring range" is an announcement. "Your jeans are the reason your outfits look cheap" is a hook. The first talks about you. The second talks about them, and makes them slightly uncomfortable. Uncomfortable gets watched.

Hooks that reliably work fall into a handful of patterns:

• Call out the audience. "If you run a trades business and your phone went quiet this winter..." The right person instantly feels seen.
• Name the pain. Say the thing they've been thinking but haven't said out loud.
• Break a belief. "Posting every day is why your Instagram isn't growing." Contradiction creates curiosity.
• Show the result first. The finished kitchen, the before and after, the product doing the thing, in the opening frame.
• Open a loop. Start a story or tease an answer the viewer has to stay to get.
• Visual interruption. Movement, an odd object, a face close to the lens, text that looks different from everything else in the feed.

On video, the hook is two things at once: what they see and what they read. Many people scroll with sound off, so the opening on-screen text has to do the work alone. Test the first frame as a still image. If it wouldn't stop you as a photo, it won't stop anyone as a video.

The Story: make them feel understood

Once you have attention, you have to earn belief. Story does that. It shows the prospect you understand their problem, often better than they do, and paints the outcome they secretly want. Facts inform. Stories move. People do not buy because they understood your features, they buy because they felt understood.

"Story" doesn't mean a three-act film. In an ad, it's a tight sequence that takes the viewer from where they are to where they want to be. The simplest structure that works:

1. The problem. Describe their situation in their own words, so specifically they wonder if you've been reading their messages.
2. The cost. What the problem is really costing them. Money, time, embarrassment, missed opportunity.
3. The turn. The insight or discovery that changes things. Why the usual fixes fail and what actually works.
4. The proof. A demonstration, a customer's words, a before and after, a number from your own business. Something that makes the claim believable.
5. The new reality. Paint the outcome. What life looks like once the problem is solved.

The storyteller matters as much as the story. A founder talking straight to camera, or a real customer explaining what changed, often beats a polished voiceover because it feels like a person, not an ad. That's why "ugly" phone videos so often outsell polished commercials. Belief is the goal, and belief comes from people.

The Offer: make saying no feel dumb

Now, and only now, you ask. A great offer is not just a product with a price. It is value stacked so high, with risk removed so completely, that saying no starts to feel like the irrational choice. Vague offers get ignored. Specific, generous, urgent offers get action.

A strong offer usually has five ingredients:

• A specific outcome. Not "marketing services" but "a new website live in 30 days."
• Stacked value. Bonuses, extras and support that make the price feel small next to what they get.
• Risk reversal. A guarantee, a trial, or terms that put the risk on you, not the buyer.
• A real reason to act now. Limited spots, a genuine deadline, a seasonal window. Never fake scarcity. People can smell it.
• One clear next step. Book, buy, claim, call. One action, stated plainly.

If your offer is "contact us for a quote," you don't have an offer. You have a door with no handle. We went deep on building offers people can't refuse in the Godfather offer.

• Hook, stop the scroll.
• Story, build the belief.
• Offer, make the ask irresistible.

A worked example: one ad, built from scratch

Say you run a Sydney cafe spending $3,000 a month on Meta ads to fill a new weekday catering service for local offices. Your current ad is a nice photo of a platter with the words "Now offering corporate catering. Enquire today." It gets likes and almost no enquiries.

Rebuild it with the formula:

• Hook: A video that opens on a sad office meeting table with a half-eaten tray of supermarket sandwiches. On-screen text: "Your team deserves better than this at the 10am meeting."
• Story: Cut to the owner in the kitchen at 6am. "Office managers told us the same thing: catering is either boring or a nightmare to organise. Late deliveries, soggy food, invoices that don't match." Quick shots of fresh platters being packed. A short clip of a real office manager saying the team actually looks forward to meetings now.
• Offer: "Order your first platter for up to 12 people this month and we'll deliver free, on time or it's on us. Book in two minutes at the link."

Same cafe, same budget, same food. The only change is the structure. The new version speaks to one specific buyer, shows it understands their problem, proves it, and makes the first order feel easy and safe. That's the whole formula in thirty seconds.

How to diagnose which part is broken

The beauty of the formula is that it makes failure diagnosable. Each part maps to numbers you can see in your ad platform:

1. Low thumb-stop and poor three-second views? Your hook is broken. Test new opening frames and first lines before you touch anything else.
2. People stop but drop off fast, with low watch time? Your story isn't landing. It's probably too slow, too generic or too much about you.
3. Good watch time but low click-through? The offer is weak or the call to action is unclear.
4. High click-through but few conversions? The landing page isn't delivering on what the ad promised. Check that the offer, wording and visuals match.

Fix one stage at a time, in order. There's no point perfecting an offer nobody sees. And if you've been blaming your targeting, read why your ROAS problem is a creative problem. More often than not, the creative is the lever.

How to test the formula without burning budget

Don't write one ad and pray. Build a simple testing system:

1. Write five hooks for the same story and offer. Hooks are the cheapest thing to test and have the biggest effect on reach.
2. Keep the story and offer steady while you find a winning hook. Change one variable at a time, or you won't know what worked.
3. Give each variation enough spend to get a real signal before you judge it. A handful of impressions tells you nothing.
4. Take the winning hook and test two or three story angles behind it.
5. Finally, test offer variations with your best hook and story combination.
6. Refresh regularly. Even winners fade as audiences see them repeatedly. Keep new hooks in the pipeline.

Keep a simple swipe file as you go. Every hook you test, the numbers it got and one line on why you think it won or lost. After a few months, you'll have your own playbook of what your specific audience responds to, which is worth more than any template on the internet.

Common mistakes that kill the formula

• Leading with the logo. Your brand name in the first second is rarely a hook for anyone who doesn't already know you.
• Clickbait hooks. A hook the story doesn't pay off burns trust and wastes the attention you earned.
• Feature dumps. A list of specs is not a story. Translate every feature into what it means for the buyer.
• Too many asks. "Follow us, visit our site, call today, and sign up to our newsletter" is four asks and zero action.
• Offer amnesia. Great hook, great story, then the ad just ends. Always tell people exactly what to do next.

Balance is everything

Most ads fail because they are all offer and no hook, or all story and no ask. Balance is everything. We engineer ads around this structure, scroll-stopping hooks, stories that land, and offers built to convert, then test relentlessly until the numbers sing.

If your ads are getting scrolled past, the fix is not a bigger budget. It is a better structure. Our advertising team writes, produces and tests campaigns built on Hook, Story, Offer for businesses that want their ad spend to actually come back. Let's build it.

Frequently asked questions

What is the hook, story, offer formula in advertising?

It's a three-part structure for persuasive ads. The hook grabs attention in the first seconds. The story shows the viewer you understand their problem and proves you can solve it. The offer gives them a specific, low-risk reason to act now. It works across video, image, audio and written ads because it follows how people naturally make buying decisions.

How long should the hook be in a video ad?

Aim to win attention within the first couple of seconds, and design the opening so it works with the sound off. That means the first frame and on-screen text should carry the hook on their own. Treat the first three seconds as the most important part of the ad, and test multiple versions of just that opening.

Does the hook, story, offer formula work for B2B ads?

Yes. B2B buyers are still people scrolling the same feeds. The hook might call out a specific role or business problem, the story might focus on cost, risk and proof from similar companies, and the offer might be a free audit, demo or guide rather than a discount. The structure stays the same, only the details change.

How do I know if my offer is strong enough?

Ask whether a sceptical buyer would feel silly saying no. A strong offer names a specific outcome, stacks enough value to make the price feel small, removes most of the risk and gives a genuine reason to act now. If your call to action is simply "contact us" or "learn more," your offer probably needs work.

It will not. Because from where your customer is standing, they cannot tell who is best. Everyone claims it. Everyone's website says "quality" and "trusted" and "results-driven." You are shouting "we are the best" into a canyon of people shouting the exact same thing. "Best" is invisible. "Different" is what gets seen.

That's the whole problem in two sentences: customers can't verify "better," so they ignore it. What they can see is a clear difference, a business that stands for something specific, serves someone specific and sounds like nobody else in the category. Get that right and excellence finally gets the credit it deserves.

The sea of sameness

Open ten websites in your industry and hide the logos. Could a stranger tell them apart? For most industries, the answer is no, same promises, same stock photos, same tired language. When everyone sounds identical, the customer has only one way left to choose: price. And competing on price is a race with no winner.

Better is a claim nobody can verify. Different is a fact nobody can ignore.

Think about how a buyer actually decides. They don't run a controlled trial of five accountants or five builders. They skim a few websites, glance at reviews, maybe ask a friend or an AI tool, and form an impression in minutes. In that window, "we deliver exceptional quality" registers as noise. Every competitor said it too. What cuts through is something the buyer can grasp instantly: "the builder who only does heritage homes," "the accountant for tradies," "the gym with no mirrors."

And it's getting harder, not easier. When AI can write polished, competent copy for everyone, competent copy stops being a signal. The only thing left to choose between is who you are. We make that case in full in when AI writes everyone's copy, brand is the only thing left to buy.

Why "best" is a trap for excellent businesses

Here's the cruel part. The better you are at the work, the more likely you are to fall into this. Excellent operators tend to believe quality speaks for itself. They pour energy into the product and treat marketing as an afterthought, a list of features and a promise to be good.

Meanwhile, a less skilled competitor with a sharp position ("the only X that does Y") takes the clients you should have won. Not because they're better. Because they're easier to understand and easier to remember.

"Best" also has a ceiling. There can only be one best, and nobody agrees who it is. "Different" has no ceiling. There's room for every business to own its own corner of the market, as long as it's willing to pick one.

Different is a decision, not an accident

• Take a real position. Stand for something specific, even if it means not being for everyone.
• Pick a fight. Name the thing your industry does wrong that you refuse to do.
• Own a niche so completely that in one corner of the market, you are the only obvious choice.
• Say what others will not. A strong point of view is magnetic, beige is forgettable.
• Build a brand with an edge, not a brand that blends into the wallpaper.

Each of these is a choice you can make on purpose. Here's what they look like in practice.

A real position sounds like "we only work with businesses doing more than a million a year" or "we only renovate bathrooms." It feels like shrinking. It's actually focusing. The right buyer hears "this is exactly for me."

A fight sounds like "we don't lock you into contracts," "we don't hide fees in the fine print," or "we'll never outsource your account." You're naming the frustration your buyers already feel with your industry and promising to be the exception.

A point of view sounds like an opinion your competitors would argue with. "Most businesses don't need a new website, they need a better offer." That kind of statement gets remembered and repeated.

How to find your different, step by step

Most businesses are already different in some way. They just haven't noticed, or they've been too cautious to say it out loud. This process helps you find it.

1. Interview your best clients. Ask five of your favourite customers why they chose you, what nearly stopped them and what they'd tell a friend. Write down their exact words. Your difference is often hiding in their answers.
2. Audit your competitors. Screenshot the headline and first paragraph of ten competitor websites. List every phrase they share. Those are the words you're banned from using.
3. List what you refuse to do. The practices you won't touch, the clients you turn away, the shortcuts you don't take. Refusals are a quick route to a clear position.
4. Find the overlap. Where does what your best clients value meet what competitors aren't saying? That gap is your position.
5. Write it in one sentence. "We are the [category] for [specific customer] who want [specific outcome], and unlike others, we [clear difference]." If it could describe a competitor, keep going.
6. Test it on real people. Say it to prospects and referral partners. Watch for recognition. If they nod politely, it's not sharp enough. If they say "oh, so you're the ones who..." you've got it.

Notice that none of this starts with a logo or a colour palette. Position first, visuals second. If you're tempted to begin with a rebrand, read why positioning comes before the logo.

Why different is safer than better

It feels risky to be polarising. It feels safe to be agreeable. But agreeable is how you become invisible, and invisible is the real risk. When you are distinctly different, you stop competing on price, you attract customers who want you specifically, and you become the category of one that no competitor can copy.

The pricing point is worth dwelling on. When you're one of many, buyers compare you line by line and ask for a discount. When you're the obvious choice for their situation, the comparison disappears and so does most of the haggling. We go deep on that in the uncomfortable psychology of premium pricing.

A worked example: two bookkeepers

Say there are two bookkeepers in the same regional town. Both are excellent. Both have happy clients.

The first one's website says: "Professional, reliable bookkeeping services for businesses of all sizes. Quality you can trust." Every enquiry starts with "how much do you charge?" and ends with the prospect getting three quotes.

The second one picks a lane. After interviewing her best clients, she notices most of them are builders and tradies who hate paperwork and fell behind on their quarterly tax obligations. So her website says: "Bookkeeping for tradies who'd rather be on the tools. We'll get your books up to date, even if you're a year behind, without the lecture." She writes about the specific headaches of trades businesses, speaks at a local builders' association and gets referred from one tradie to the next.

She's not better at bookkeeping than the first one. But when a builder asks a mate "who does your books?", there's only one obvious answer. She's easier to remember, easier to refer and far less likely to be compared on price. (The two businesses are hypothetical, but the pattern plays out in every industry.)

The mistakes that keep businesses beige

• Hedging the niche. "We specialise in tradies, but we work with everyone." The second half cancels the first.
• Choosing a difference nobody cares about. Being the only firm with a purple logo is different, but it's not a reason to buy. Your difference has to matter to the buyer.
• Saying it once. A position only works if it shows up everywhere, on your site, in your proposals, in how your team answers the phone.
• Letting the committee sand it down. Every round of internal feedback removes a little edge. Protect the sharp version.
• Blaming the market. If leads are slow and price pressure is high, the problem is usually the message, not the competition. More on that in your competitors aren't the problem, your message is.

How to roll it out and know it's working

Finding your position is half the job. The other half is making sure every customer touchpoint says it. Start with the places buyers see first:

1. Your homepage headline. Replace the generic promise with your one-sentence position. This single change often does more than a full redesign.
2. Your Google Business Profile and social bios. Same words, same position. Consistency is what makes a difference stick.
3. Your proposals and quotes. Open with why you're different for this client, not with a company history.
4. Your team's elevator pitch. If three staff describe the business three different ways, the market hears confusion.

Then watch for the signals. Are enquiries coming from the kind of customer you described? Are prospects repeating your language back to you ("we heard you're the ones who...")? Are fewer conversations starting with price? Is your close rate improving, even if total enquiry numbers dip at first? A good position often brings fewer tyre-kickers and more buyers, which is exactly the trade you want. Give it a few months before you judge it, because recognition builds slowly and then all at once.

Excellence deserves to be seen

We help ambitious brands trade "better" for "unmistakable", finding the position, the point of view, and the edge that makes you the only logical choice for the right people. Our consulting work starts exactly where this article does: with your best clients, your competitors and the truth about what makes you different.

If your marketing sounds like everyone else's, being excellent will not save you. Being different will. Let's find your different.

Frequently asked questions

How do I make my business stand out from competitors?

Stop leading with claims of quality and start leading with a clear difference. Talk to your best clients about why they chose you, study what every competitor says, and find the gap between the two. Then commit to one specific position, for one specific customer, and repeat it everywhere. Standing out is a decision to be distinct, not a louder version of the same promise.

Is it risky to focus on a niche?

It feels risky, but staying generic is usually riskier. A niche makes you easier to understand, remember and refer, which tends to bring more of the right enquiries, not fewer. You can still accept work outside your niche. You simply stop marketing to everyone, so your message lands hard with the people most likely to buy and pay well.

What is a unique value proposition, in plain English?

A unique value proposition is a short statement of who you serve, what outcome you deliver and why you're different from the alternatives. A strong one is specific enough that it couldn't describe a competitor. If yours contains words like "quality," "trusted" or "tailored solutions" without anything more concrete, it probably isn't doing its job yet.

Why do customers always choose the cheapest option?

Customers choose on price when they can't see any other meaningful difference between options. If every business looks and sounds the same, price becomes the only thing left to compare. Give buyers a clear, relevant reason to choose you specifically, and price drops down the list. Differentiation is one of the most reliable ways to escape a discount war.

Why does your podcast sound amateur? Almost always it's the room, the way you use the mic and the lack of a proper mix, not the microphone itself. Fix those three and a modest setup can sound broadcast-ready, and most of it can be sorted in a single weekend.

Here's what's actually going wrong, and a step-by-step plan to fix it.

It is not the microphone

Most people think professional sound is a gear problem, so they buy an expensive microphone and stay confused when it still sounds off. The mic is a small part of it. Most of it is the room, the technique, and what happens after you hit record. A cheap mic used well beats an expensive mic used badly, every time.

In fact, an expensive studio condenser can make things worse. Sensitive mics hear everything: the fridge, the traffic, the echo bouncing off your bare walls. Put one in an untreated home office and you've bought a very accurate recording of a bad room.

For most spoken-word shows recorded at home or in an office, a decent dynamic microphone is the smarter starting point. Dynamics pick up the voice in front of them and reject more of what's around them. They're forgiving. And forgiving is exactly what you want while you're learning.

Nobody remembers your best episode if they could not stand to listen to it. Sound is the price of admission.

What actually makes it sound pro

• The room. Echo is the number one giveaway. Soft surfaces and a treated space beat any microphone upgrade.
• Mic technique. Consistent distance and position matter more than the badge on the mic.
• Levels. Recording too quiet or clipping too hot ruins audio before editing even begins.
• The mix. EQ, compression, and noise reduction are where amateur becomes professional.
• Consistency. Every episode landing at the same polished, even loudness builds trust.

Notice the order. Problems at the top of the list can't be fully fixed by anything lower down. Software can reduce echo, but it can't remove it cleanly. It can lift a quiet recording, but it lifts the noise with it. It can't rebuild audio that clipped. Get the capture right and post-production becomes polishing instead of damage control.

Fix the room first

Clap your hands once in the room you record in. If you hear a ring or a flutter after the clap, your listeners hear that on every word. That's reflected sound, and it's the single biggest reason podcasts sound like they were recorded in a bathroom.

You don't need a studio build to fix it. You need soft, absorbent stuff between your voice and hard surfaces.

• Pick the right room. A small bedroom full of furniture, carpet and curtains usually beats a big, empty, stylish room with polished floors and glass.
• Add soft surfaces. Rugs, heavy curtains, a couch, a bookshelf full of books. Even hanging a thick blanket behind you and in front of you makes an obvious difference.
• Get off the wall. Don't sit with your face a short distance from a bare wall. Sound bounces straight back into the mic.
• Kill the noise sources. Air conditioning, fans, fridges, a humming computer. Turn off what you can and move the mic away from what you can't.

Acoustic panels are worth it if you record every week in the same space. But try the free fixes first. They get you most of the way.

Mic technique and levels

The best room in the world won't save you if your distance from the mic keeps changing. Every time you lean back, your voice gets thinner and the room gets louder. Every time you lean in, the bass swells. Listeners hear that as the sound "wobbling."

1. Find your distance and hold it. For most dynamic mics, roughly a fist's width away works well. Use a boom arm so the mic stays put and you don't hunch over a desk stand.
2. Speak slightly across the mic, not straight into it. This softens the pops on P and B sounds. Add a pop filter or foam windscreen too.
3. Set levels with your real voice. Do a test at the energy you'll actually use, including when you laugh. Aim for healthy peaks that never touch the top of the meter.
4. Record each person on their own track. Separate tracks let you fix one speaker without wrecking the other. With remote guests, have each person record locally where possible rather than relying on the call audio.
5. Record a few seconds of silence. That room tone is gold later for noise reduction and smoothing edits.

The invisible craft

Great audio is invisible, you only notice it when it is missing. That polished, effortless sound you hear on top shows is not luck or expensive toys. It is a hundred small, deliberate decisions in capture and post-production that most creators never learn to make. That is exactly the gap between "content that happens to be recorded" and "a show people trust."

Here's what that post-production chain usually involves, in plain English:

• Cleanup. Gentle noise reduction, removing mouth clicks, and cutting long pauses, false starts and crosstalk.
• EQ. Rolling off the low rumble, taming harshness and nasal tones, and adding a touch of clarity so voices cut through on phone speakers and earbuds.
• Compression. Evening out the gap between loud and quiet moments so listeners aren't riding the volume button while driving.
• Balancing voices. Making host and guest sit at the same perceived level.
• Loudness and mastering. Bringing the whole episode to a consistent loudness so it sits comfortably next to the other shows in your listener's queue.

Then do the test most creators skip. Listen back on the devices your audience actually uses: cheap earbuds, a phone speaker, a car stereo and a laptop. A mix that sounds lush on studio headphones can turn muddy in a car or vanish on a phone. If voices stay clear and even across all of them, you're done. If one device exposes a problem, fix it before you publish, because a good share of your listeners will hear exactly that version.

If the last step sounds like a music thing, it is, and the logic carries straight over. We explain the difference in mixing versus mastering if you want to know what each stage does.

A worked example: the weekend fix

Here's a hypothetical. Say you host a business interview show from your home office. Hard floors, a big window, the mic on a desk stand, and guests joining over a video call. Listeners say the content is great but it's "a bit echoey," and completion rates are sliding.

Saturday morning: you move recording into the carpeted spare room, hang a heavy blanket on the wall behind the desk and throw a rug over the floorboards. You clap-test and the ring is noticeably shorter.

Saturday afternoon: you mount the mic on a boom arm, add a foam cover, mark your seated position and record a test at full energy. You adjust gain until your loudest laugh stays clear of the top of the meter.

Sunday: you set up your recording platform so each guest records their own local track, and you write a one-page guest checklist: wired headphones, quiet room, laptop off the fan, sit close to the mic. Then you build a simple processing template so every episode goes through the same cleanup, EQ, compression and loudness steps.

None of that required new expensive gear. The next episode sounds like a different show. That's the point: most of the jump from amateur to pro is process, not purchases.

Common mistakes that scream amateur

• Recording in the nicest-looking room. Great on camera, terrible for sound. If you film video podcasts, treat the areas out of frame.
• Trusting headphones you never wear. Monitor with headphones while recording so you catch problems live, not in the edit.
• Over-processing. Heavy noise reduction makes voices sound underwater and robotic. Clean capture needs less fixing.
• Mismatched guest audio. A crisp host next to a laptop-mic guest makes the whole show sound cheap. Send guests a checklist, or a basic mic.
• Inconsistent loudness between episodes. If listeners have to change the volume every week, you're adding friction to a habit you want them to keep.
• Ignoring intro music quality. A cheap, generic sting sets the tone before you say a word. Our take on why royalty-free music can quietly hurt your brand applies here too.

When to fix it yourself and when to hand it off

Do the room, the technique and the levels yourself. Nobody else can sit in your chair. That's your weekend.

Post-production is where the decision gets real. Learning to edit, mix and master well takes time, and that time comes directly out of booking guests, promoting episodes and running your business. If your podcast is part of how you win clients, the sound is part of your brand. We've written about how sonic branding shapes how people feel about a business, and your show's sound is one of the most-heard pieces of it.

A good rule: if an episode takes you longer to edit than to record and it still doesn't sound like the shows you admire, it's time to hand it off.

Give your content the sound it deserves

We take raw recordings and make them sound like the shows people actually finish, clean, warm, consistent, professional. If your content deserves better than the sound it is trapped in, let's fix the thing your listeners judge first. Our music and sound team handles podcast editing, mixing and mastering, so you can get back to making the show.

Frequently asked questions

What is the best microphone for a beginner podcast?

For most people recording at home or in an office, a good dynamic microphone is the safest choice. Dynamics focus on the voice right in front of them and pick up less room echo and background noise than sensitive condenser mics. Pair it with a boom arm and a pop filter, and treat the room before you spend more on gear.

How do I get rid of echo in my podcast recordings?

Stop it at the source. Record in a smaller room with carpet, curtains and soft furniture, hang blankets or panels behind and in front of you, and keep a consistent, close distance to the mic. Software can reduce echo afterwards, but it can't remove it cleanly, and heavy processing often leaves voices sounding thin or robotic.

Why do my remote podcast guests sound bad?

Usually because their audio is coming through the video call's compressed stream, recorded on a laptop mic in an echoey room. Use a platform that records each person locally, send guests a short setup checklist covering headphones, a quiet room and mic position, and consider posting a basic mic to regular guests.

Is it worth paying someone to edit my podcast?

If the podcast supports your business and editing is eating hours you'd rather spend on guests, promotion or clients, usually yes. A professional editor brings consistent cleanup, mixing and loudness to every episode, which keeps listeners coming back. Handle the room and recording technique yourself, then let a specialist handle post-production.

The brands that always seem to have fresh content are not filming constantly. They are multiplying, extracting months of content from a single, well-planned shoot.

So can one video shoot really give you 90 days of content? Yes, if you plan it for multiplication before anyone picks up a camera. One focused production day, shot with every format and platform in mind, can produce hero videos, dozens of short clips, stills, quotes and teasers, enough to post several times a week for three months. The difference isn't budget. It's the plan.

Shoot once, plan for a hundred pieces

The secret is in the intention. An amateur films a video. A pro films a video knowing it will become dozens of assets: the hero cut, the short vertical clips, the quotes, the behind-the-scenes, the stills, the teasers. Same day on set, same cost, but planned for multiplication, the output is many times larger.

The most expensive part of content is not making it. It is wasting it. Most brands throw away most of what they paid to film.

Think about where the money actually goes on a shoot day. The crew, the gear, the location, the talent, the travel, the hours spent getting everyone in the same room. Those costs are almost identical whether you walk away with one video or eighty assets. Every extra piece you pull from the same day drives your cost per asset down. Waste is the only real overspend.

That's also why the "film a little every week" approach burns so many founders out. Every small shoot carries its own setup, its own planning, its own lost afternoon. You pay the overhead fifty times instead of once.

Plan before you press record

Multiplication is decided in pre-production, not the edit suite. If it isn't captured on the day, no editor can create it later. Here's the planning sequence that makes a shoot multiply.

1. Start with the calendar, not the camera. Map the next 90 days: launches, offers, seasonal moments, the questions customers ask most. That list decides what you need to film.
2. Pick three to five content pillars. For example: how it works, who it's for, behind the scenes, customer proof, founder opinion. Every shot on the list should serve one pillar.
3. Write a shot list that covers every format. Horizontal for the website and YouTube, vertical for Reels, TikTok and Shorts, square-safe framing for feeds. Shoot wide enough to crop both ways, or shoot key moments twice.
4. Script the talking points, not the words. Prepare a list of short, standalone answers the founder or spokesperson can deliver in 20 to 60 seconds each. Each one becomes its own clip.
5. Schedule stills time. Block time for a photographer, or for pulling high-resolution frames, so you walk away with images, not just video.
6. Plan the b-roll bank. Hands at work, product details, the space, the team, the process. B-roll is the connective tissue that turns talking heads into watchable edits for months.

How one shoot becomes ninety days

• Capture with repurposing in mind, multiple angles, formats, and moments deliberately banked on the day.
• Cut the pillars first, the long-form hero pieces that anchor everything.
• Slice into shorts, dozens of vertical clips engineered to stop the scroll.
• Pull the stills, frames become photos, thumbnails, and graphics.
• Schedule the drip, release it steadily so you look consistently present, not sporadically busy.

Here's how that breaks down in practice. A well-planned shoot day might yield:

• One or two hero pieces: a brand film or explainer for your homepage, a long-form interview for YouTube or a podcast feed.
• Twenty to forty short clips: each standalone answer, each strong moment, each mini-story cut vertically with captions.
• Quote graphics and text posts: the best lines from the interview, pulled into carousels and email content.
• A stills library: portraits, product shots and detail frames for social, the website, ads and press.
• Behind-the-scenes and teasers: the setup, the bloopers, the "coming soon" moments that build anticipation for the hero piece.
• Ad creative: multiple hooks cut from the same footage, so you can test angles without reshooting.

Spread across three months, that's more than enough to post consistently on your main channels without ever scrambling.

A worked example: one day, one quarter

Here's a hypothetical. Say you run a boutique Gold Coast gym and you book one production day before the new year rush.

In the morning you film a two minute brand film: the coaches, the classes, the community, finished with a founder line about why the gym exists. Then you sit the founder down for an hour and ask twenty questions members always ask. "I'm unfit, is this for me?" "How often should I train?" "What should I eat before a class?" Each answer is 30 to 60 seconds, delivered straight to camera.

In the afternoon you film three real members talking about why they joined (with their permission), grab b-roll of every class type, and shoot portraits of each coach. A photographer pulls stills of the space and the equipment.

Now look at the output. The brand film goes on the homepage and becomes your ad anchor. Twenty FAQ answers become twenty Reels, each one a different entry point for a different kind of nervous beginner. Three member stories become proof-driven ads. The coach portraits feed the website and a "meet the coach" series. The b-roll gets reused under captions and voiceovers all quarter. That's two to three posts a week, plus ad creative, from a single day. No weekly scramble, no founder filming in the car park at 6am.

What it takes, and what it costs

A multiplication shoot costs more than a quick video and far less than filming every week. The real investment is in three places.

Planning time. Expect the pre-production to take longer than the shoot itself. Content mapping, shot lists and talking points are where the value is created. Skip them and you're back to a single-use shoot.

A crew that thinks in formats. Someone on set needs to be watching framing for both vertical and horizontal, catching extra angles, and ticking off the shot list. That's a different mindset from filming one commercial.

Editing capacity. This is the part most brands underestimate. Forty clips need forty edits, each with captions, a hook in the first seconds and a clear ending. Budget for the edit, or the footage sits on a hard drive, which is exactly the trap we started with.

The payoff is the maths. Divide the total cost by the number of usable assets and compare it with what you're paying per piece now. For most brands that currently shoot one-off videos, the multiplication model wins by a wide margin.

If you're working with a smaller budget, the model still holds. Scale it down rather than skipping it. A half day, one camera, one presenter, a tight list of fifteen questions and a phone for behind-the-scenes clips will still outproduce a month of ad hoc filming. The principle is the same at every size: plan everything, capture everything, waste nothing.

Mistakes that kill the multiplication

• Only shooting horizontal. If you frame everything for a widescreen hero film, you'll be cropping heads off in every vertical edit. Plan both from the start.
• Long rambling answers. A five minute answer is one clip at best. A tight 40 second answer is a finished post. Coach your talent to answer in short, complete thoughts.
• No hooks. Short-form lives or dies on the first second or two. Plan opening lines and strong first frames on the day, and cut thumbnails with intent. We cover that in the psychology of the scroll-stopping thumbnail.
• Over-polishing everything. Not every piece needs a cinematic grade. Some of the best-performing clips are raw behind-the-scenes moments, which is why "ugly" phone videos can outsell a polished commercial. Capture both.
• Dumping it all at once. Posting twenty clips in the first week wastes the library. Drip it out on a schedule.
• No filing system. Label and log every clip by pillar and topic. If your team can't find the right clip in two minutes, they'll shoot a new one.

Presence without the burnout

This is how you go from a content emergency every week to a calm, full pipeline, how a brand starts to feel omnipresent without a full-time crew or a founder chained to a phone. One focused day of production, engineered properly, feeds you for a season.

Consistency is what builds recognition. Being seen repeatedly, across platforms, with the same face and the same message, is how a small brand starts to feel bigger than it is. We break down the whole strategy in how to be everywhere on a startup budget. A multiplication shoot is the fuel for that engine.

To measure it, track three things across the quarter: how many weeks you posted on schedule, which pillars and hooks earned the most watch time and saves, and what the content did for enquiries or sales. The winners tell you exactly what to shoot more of next time, so every production day gets smarter than the last.

Build a smarter shoot, not more of them

We plan and shoot for multiplication, turning a single production day into a library of scroll-stopping assets that keep you visible for months. If you are filming constantly and still running dry, you do not need more shoots. You need a smarter one. Let's build it.

That's the core of how our photo and video team works at LIQID NOISE: strategy and shot list first, one well-run day on set, then an edit plan that keeps your channels full long after the crew has packed up.

Frequently asked questions

How much content can you get from one video shoot?

It depends on planning more than budget. A shoot designed for repurposing, with a detailed shot list, standalone talking points, b-roll and stills time, can produce one or two hero videos, dozens of short vertical clips, quote graphics and a photo library. A shoot planned around a single video usually produces that one video and little else.

How do you repurpose long videos into short clips?

Start by finding standalone moments: answers, stories or strong lines that make sense without context. Cut each into a vertical clip with a hook in the first seconds, captions for silent viewing and a clear ending. It works far better when talking points were scripted as short, complete answers on the day, rather than hunting through long rambling takes later.

Should I film content for vertical or horizontal?

Both, planned from the start. Horizontal suits websites, YouTube and brand films. Vertical suits Reels, TikTok and Shorts. Either frame wide with enough room to crop for both, or shoot key moments twice in each orientation. Deciding this before the shoot saves you from awkward crops and unusable footage later.

How often should a small business post video content?

Consistency beats volume. Pick a rhythm you can sustain, such as two or three posts a week on your main platform, and hold it for months. A single planned shoot can cover that schedule for a quarter. Posting daily for two weeks and then disappearing does less for your brand than a steady rhythm you never break.

So if your return on ad spend is sliding, here's the blunt answer: the fix is almost never a new audience. It's a better ad. On Meta, TikTok and YouTube today, the creative does the targeting. The hook decides who stops, the message decides who stays, and the algorithm learns who to find more of from the people your ad attracts.

That's good news, by the way. You can't out-engineer the platform's own machine learning at its own game. But you can absolutely out-create your competitors.

Creative is the new targeting

When everyone has access to the same auto-optimising delivery, the only real lever left is the creative itself. The ad that stops the scroll wins the auction. Full stop.

Think about how delivery actually works now. You give the platform a broad audience and an ad. It shows the ad to a spread of people, watches who engages, clicks and buys, then goes looking for more people who behave like them. The ad is the filter. An ad that speaks directly to busy mums who hate meal prep will find busy mums who hate meal prep, whether you ticked a "parents" box or not.

That's why two advertisers with identical settings, identical budgets and identical products can get wildly different results. The settings are the same. The signal is different. If you want the full case for letting go of the audience builder, read why broad targeting wins in 2026.

You no longer beat the algorithm with settings. You beat it with a better first three seconds.

How to tell if you have a creative problem

Before you tear anything down, diagnose it. Creative problems leave fingerprints in your ad account. Look for these:

• People aren't stopping. Low thumb-stop or three-second view rates mean your opening isn't earning attention. The rest of the ad never gets a chance.
• People stop but don't stay. Decent early views that drop off a cliff mean the hook promised something the middle didn't deliver.
• People watch but don't click. Strong watch time with a weak click-through rate means the ad entertains but never gives a reason to act.
• People click but don't buy. That's often a landing page or offer problem, not a creative one. Check the page matches what the ad promised before you blame the video.
• Everything was fine, then it wasn't. Results that were strong and slowly decayed point to fatigue. The audience has seen it. You need fresh angles, not new settings.

Work top down. There's no point polishing the call to action on an ad nobody watches past the first second.

The anatomy of a high-ROAS ad

• The hook (first 3 seconds): Pattern interrupt. A bold claim, a visual surprise, or a problem stated out loud before the viewer can swipe.
• The tension: Agitate the pain or amplify the desire. Make staying feel more interesting than leaving.
• The proof: Show, don't tell. Demonstrations, results, and real reactions outperform claims every time.
• The ask: One clear action. Confusion kills conversion.

Here's what each part looks like when it's done well.

Hooks that work usually do one of four things: call out the viewer ("If you own a café, stop scrolling"), name a painful problem ("Your oven is costing you $40 a week"), make a surprising claim ("We stopped running Google Ads and sales went up"), or show something visually odd enough that the brain has to look. The hook should be both spoken and on screen, since plenty of people watch on mute.

Tension is the reason to keep watching. Open a loop: "Here's the mistake almost every renovator makes." Now the viewer needs the answer.

Proof is where most ads are weakest. A founder saying "our product is amazing" proves nothing. A customer showing the before and after proves everything. Demonstrations beat descriptions.

The ask should be single and specific. "Tap below to book your free measure and quote" beats "Learn more about our range of services." If you want a deeper version of this structure, the hook, story, offer formula breaks it down further.

Volume + iteration = control

The brands winning on paid media aren't producing one perfect ad. They're shipping dozens of variations, letting the data crown the winners, and reinvesting into new angles. We treat creative like a portfolio, constantly testing hooks, formats, and offers so your cost-per-acquisition trends down while spend scales up.

But volume doesn't mean fifty near-identical ads with a different background colour. Platforms have become better at recognising when ads are basically the same, and near-duplicates tend to compete with each other rather than reach new people. What you want is genuine variety in the things that matter:

• Angles. Different reasons to buy. Saving time, saving money, looking good, avoiding a disaster.
• Formats. Talking head, customer testimonial, product demo, static image, carousel, founder story.
• Hooks. Several openings on the same body of the ad, because the first three seconds carry most of the weight.
• Offers. A free quote versus a discount versus a guarantee can shift results more than any edit.

Meta's own delivery changes have pushed hard in this direction. We unpacked why in Meta's Andromeda update, where genuinely different concepts beat piles of lookalike variants.

A simple creative testing system you can run this month

You don't need a big budget to test properly. You need a structure, and the discipline to stick with it.

1. Write down three angles. Three genuinely different reasons someone would buy from you. Pull them from customer reviews and sales calls, not from a brainstorm.
2. Make two formats per angle. For example, one talking-head video and one static image. That's six concepts.
3. Write three hooks for each video. Same body, different first three seconds.
4. Launch them in one broad campaign. Let the platform distribute spend. Resist the urge to switch things off in the first few days.
5. Judge on cost per purchase or cost per qualified lead. Not likes, not clicks. Give each concept enough spend to tell you something before you decide.
6. Kill the losers, clone the winners. Take the winning angle and make three new variations of it. Take the winning hook and try it on a different format.
7. Repeat every two to four weeks. Winning ads wear out. A steady flow of new concepts keeps your costs from creeping up.

That rhythm matters more than any single ad. If you're wondering why your winners seem to die faster than they used to, read about ad fatigue and the two-week lifespan.

Producing creative in volume without a film crew

The obvious objection: "I can't afford to make twenty ads a month." You don't need to. Most of the best-performing ads in small accounts are cheap to make. A founder talking to a phone camera about the one mistake customers make. A customer answering a single question on video. A screen recording of the product doing its job. A plain static with a strong headline and a real review.

The trick is to batch it. Spend one morning recording ten short pieces to camera, each built around a different angle or hook. Pair that with real customer footage and a few statics, and you have a month of testing material. Save the polished production budget for the concepts that have already proven they sell.

A worked example: the kitchen renovator

Say you run a Melbourne kitchen renovation business spending $5,000 a month on Meta ads. You have one ad running: a polished slideshow of finished kitchens with the headline "Quality Kitchens, Affordable Prices." Results were decent a few months ago. Now leads are costing more every week, and you've tried three new interest audiences with no change.

That's a creative problem. The ad says what every competitor says, and the audience has seen it.

So you build three angles. Angle one: "We finish on time, or you don't pay the final invoice" (a guarantee, if you can genuinely offer it). Angle two: "The three mistakes that blow out a kitchen budget" (education). Angle three: a customer walking through their new kitchen explaining what they were nervous about before they started (proof).

Each gets a video and a static. The videos get three hooks each. Within a few weeks, the spend data tells you which angle the market cares about. Maybe the budget mistakes video pulls the cheapest qualified leads by a clear margin. Now you know your buyers are worried about cost blowouts, and every new ad leans into that. Same budget. Same broad audience. Very different result. (Every number here is hypothetical, but the pattern is how testing works.)

Common creative mistakes that crush ROAS

• Leading with the logo. Nobody cares who you are in second one. They care whether this is relevant to them.
• Making ads that look like ads. Heavy branding and stock footage get scrolled past. Native-looking content earns a second look.
• Talking about features instead of outcomes. "Soft-close drawers" is a feature. "No more slamming at 6am while the kids sleep" is an outcome.
• Testing tiny changes. A new font won't teach you anything. A new angle will.
• Turning ads off too early. Judge on enough data to mean something, not on one bad day.
• Letting winners run forever. Every ad has a shelf life. Have the next one ready before this one fades.

Where to start this week

Open your ad account. Look at your top three ads by spend. For each one, ask: what's the hook, what's the angle, and what's the proof? If you can't answer clearly, neither can your audience.

Then read your last twenty customer reviews and write down the exact words people use to describe why they bought. Those phrases are your next hooks. Customers write better ad copy than most marketers, because they describe the problem in the language other buyers actually use.

At LIQID NOISE, our advertising work is built on this: strategy-led creative, produced in volume, tested with discipline and judged on revenue. If your targeting is already broad and your ROAS still won't move, it's time to stop fiddling with settings and start fixing the ads.

Frequently asked questions

Does ad targeting still matter on Meta and TikTok?

Targeting still matters at the edges. You should set the right country, exclude existing customers where it makes sense and give the platform accurate conversion data. But for most advertisers, detailed interest stacking adds little. The algorithm finds buyers based on who responds to your ad, so the creative now does most of the work that manual targeting used to do.

How many ad creatives should I test at once?

For most small to medium accounts, start with three genuinely different angles and two formats each, around six concepts, with a few hook variations on the videos. That's enough variety to learn something without spreading a modest budget too thin. As spend grows, add more concepts. Focus on real differences in angle and format, not small tweaks.

Why did my ROAS suddenly drop?

Sudden ROAS drops usually come from creative fatigue, a tracking problem, a landing page change or a shift in competition. Check tracking first, since broken purchase events make results look worse than they are. Then look at frequency and whether your best ads have been running for weeks. If people have seen them repeatedly, fresh creative is usually the fix.

What is a good hook for a video ad?

A good hook makes the right person stop in the first three seconds. It might call out the viewer directly, name a painful problem, make a surprising claim or show something visually unexpected. It should appear as on-screen text as well as speech, since many people watch on mute. Test several hooks on the same ad body to find the one that works.

So can a book actually bring you clients? Yes, if you write it for that job from page one. A deal-closing book is written for your ideal buyer, solves a problem they already have, shows how you think, and gives them a clear next step. Then you put it in front of prospects on purpose, before and during the sale, instead of waiting for strangers to find it on Amazon.

Authority you cannot fake

Anyone can call themselves an expert. A book is different, it is proof. When you hand a prospect your book, or they find it before they find you, the entire relationship changes. You are no longer a vendor pitching for attention. You are the author, the authority, the person who literally wrote the book on this. That shift does more for your close rate than a hundred sales calls.

A business card gets thrown away. A book gets kept, read, and remembered, and it sells for you long after the meeting ends.

Why does a book carry that weight when a blog post or a brochure doesn't? Because everyone knows what it takes. Writing a book means organising your thinking into a complete argument, standing behind it in print and putting your name on the cover. A prospect doesn't need to read every page to feel that. The object itself says: this person has done the work.

And for the ones who do read it, something even more valuable happens. They spend a few hours alone with your ideas, in your voice, at their own pace. No sales call gets that much undivided attention. By the time they contact you, they already understand your approach, agree with your diagnosis and want your help to apply it. We unpack that idea further in your book is the most powerful business card you'll ever print.

How a book quietly generates business

• It pre-sells. Prospects arrive already trusting you, already convinced, ready to buy.
• It filters. The people who read it and reach out are warmer and better-fit than any cold lead.
• It opens doors, stages, podcasts, media, and partnerships that "expert" alone never gets you into.
• It works around the clock, making your case while you sleep, travel, or focus elsewhere.
• It compounds, write it once, and it keeps building authority for years.

Notice what's missing from that list: book sales. For most founders and consultants, royalties are a rounding error. The money is in the clients, speaking fees and partnerships the book brings in. That changes every decision you make about it, from the topic to the price to how many copies you give away.

Trophy book vs deal-closing book

The difference is mostly decided before a word gets written. Here's how the two compare.

• Audience. A trophy book is written for "everyone interested in business." A deal-closing book is written for one specific buyer, like owners of trades businesses doing a few million a year, or HR directors at mid-size firms.
• Topic. A trophy book tells your life story. A deal-closing book solves the exact problem your service solves, and your story shows up only where it serves the lesson.
• Method. A trophy book shares scattered tips. A deal-closing book lays out your method, named and step by step, so the reader can see how you work and how much there is to get right.
• Ending. A trophy book ends with "thanks for reading." A deal-closing book ends with an invitation, and references a free resource, assessment or conversation throughout.
• Distribution. A trophy book gets a launch party and a shelf. A deal-closing book gets built into the sales process.

That last point surprises people. Giving your method away in print feels like it would cost you clients. In practice, it does the opposite. Most readers don't want to do it themselves. They want the person who clearly knows how to do it for them.

How to plan a book that closes deals

1. Define the reader in one sentence. Who are they, what are they struggling with, and what have they already tried? If you can't write this sentence, you're not ready to outline.
2. Name the transformation. Where is the reader at the start of the book, and where will they be at the end? That gap is your book.
3. Map your method into chapters. Most experts already have a process they use with clients. Name the stages and give each one a chapter. This becomes the spine of the book.
4. Collect your stories and examples. Every chapter needs at least one story that makes the idea real. Use your own experiences, and get permission before sharing anything about a client.
5. Plan the bridge to your business. Decide what the reader should do next. A scorecard, a template, a workshop or a strategy call. Reference it naturally where it helps, not on every page.
6. Choose your publishing path. For most business owners, self-publishing or hybrid publishing gives more control, faster timelines and full rights to give copies away. Our guide to self-publishing versus traditional lays out the trade-offs.

And keep it readable. A business book that reads like a manual won't get finished, and an unfinished book doesn't close anything. Short chapters, clear headings, real stories and a point of view on every page. More on that in how to write a business book people actually finish.

The asset that pays forever

Think about the leverage. Every other marketing effort demands constant fuel, more ads, more posts, more outreach. A book is written once and then works indefinitely, turning your expertise into an asset that builds trust and generates leads long after the writing is done. Few investments in a business return for as long.

It also feeds everything else you do. Every chapter can become a keynote, a workshop, a podcast talking point, an article series, a video series and an email sequence. The thinking is done. You're just repackaging it for different rooms.

How to put the book to work in your sales process

This is where most authors drop the ball. They publish, post about it for a week and wait. A deal-closing book needs a job description.

• Before the first call. Send a copy (physical if the deal is big enough) as soon as a qualified prospect books a meeting. They arrive having read a chapter or two, and the conversation starts further along.
• After the first call. Mail a signed copy with a handwritten note that references the chapter most relevant to what they told you. Few follow-ups are that memorable.
• As a lead magnet. Offer a free digital chapter or the full ebook in exchange for an email, then follow up with a short sequence that expands on the ideas.
• At events. Give copies to attendees when you speak. The talk gets them interested, the book keeps you in their office for months.
• For referral partners. Give a box of books to accountants, lawyers and advisers who serve the same clients. It's easier to refer someone when you can hand over their book.

A worked example: the HR consultant

Say you run a small HR consultancy in Sydney that helps growing businesses fix high staff turnover. Your sales cycle is long. Prospects take three or four meetings to trust you, and many go quiet in between.

You write a short, focused book: a named five-stage method for cutting turnover, one chapter per stage, each with a story from your years of practice (anonymised and with permission). It ends with an invitation to a free turnover assessment.

Now the process changes. When a prospect books a first meeting, they get the book in the mail with a note. Many arrive having read the first chapters, already using your language for their problem. The meetings needed to build trust drop, because the book has done some of that work. Referral partners start handing it out. And when you speak at an industry event, you leave copies on every seat. (This scenario is hypothetical, but it's exactly how a book should be deployed.)

What it takes, and the mistakes to avoid

A focused business book doesn't need to be huge. A tight book the reader can finish in a couple of sittings usually does the job better than a doorstop. The time commitment depends on how you write it. Writing alone, around a busy business, can drag on for years. Working with a ghostwriter or book coach who interviews you and shapes your thinking into chapters can get it done in months. We laid out a realistic timeline in the 90-day book for busy founders.

The mistakes that stop a book from closing deals:

• Writing for everyone. A book for everyone persuades no one. Narrow the reader.
• Holding back the good stuff. Vague books build no trust. Show your real method.
• Skipping the bridge. If the reader finishes and doesn't know what to do next, you've written a trophy.
• Treating the launch as the finish line. The book's real work starts after launch, inside your sales process.
• Cutting corners on editing and design. A book is proof of quality. A sloppy one proves the opposite.

Your most powerful trust-builder, unbuilt

We help experts and founders turn what they know into a book that builds authority and quietly fills the pipeline, from the idea to the finished, published asset. Through our books and publishing work, that means shaping the idea, writing it with you, and producing a book that looks and reads like it belongs next to the best in your field.

If you have a business and a message but no book, you are leaving your most powerful trust-builder unbuilt. Let's change that.

Frequently asked questions

Can writing a book really get you more clients?

Yes, when it's written for your ideal buyer and used deliberately in your sales process. A focused business book builds trust before the first meeting, filters in better-fit prospects and gives referral partners something easy to pass on. The clients come from how you use the book, not from bookstore sales, so plan its distribution as carefully as its content.

How long should a business book be?

Long enough to solve one clear problem properly, and short enough that a busy reader can finish it. For most lead-generating business books, that means a tight, focused read that can be finished in a couple of sittings. Cut anything that doesn't serve the reader's transformation. A shorter book that gets finished does far more for your business than a long one that sits unread.

Should a business book give away your whole method?

Mostly, yes. A book that hides the good stuff reads like a sales brochure and builds little trust. Sharing your real method shows depth and helps readers see how much there is to get right. Most readers would still rather hire the person who clearly knows how to do it than try it alone, which is exactly why generous books close deals.

Do I need a traditional publisher for a book to build authority?

No. For business owners, authority comes from the quality of the ideas, the editing and the design, and from how you use the book. Self-publishing or hybrid publishing usually offers more control, faster timelines and the freedom to give copies away. A traditional deal can add prestige, but it isn't required for a book to win clients.

Why do WebGL sites convert better? Because they turn a visitor from a skimmer into a participant. A 3D scene that responds to the cursor, the scroll and the touch holds attention longer, lets people inspect a product from every angle, and answers questions a flat photo can't. Built properly, it does all that without slowing the page down. Built badly, it's an expensive way to lose customers.

This is the guide to building it properly.

Attention is the only currency that matters

A 2006 study by Lindgaard and colleagues found people judge a web page's visual appeal within about 50 milliseconds of seeing it. A static hero image gives them nothing to explore. A reactive 3D scene, one that responds to their cursor, scroll, and motion, creates a feedback loop. They move, the world moves, and suddenly they are playing instead of skimming.

That shift from passive to active is the whole point. A reader skims. A player explores. And exploration is what buyers do right before they commit, turning the thing over, zooming in, checking the back, imagining it on their desk.

A flat page asks the visitor to imagine all of that. A 3D page lets them do it.

Engagement isn't a vanity metric. Every extra second on page is a step closer to conversion.

What WebGL actually is, in plain English

WebGL is the technology that lets a browser draw real-time 3D graphics using the device's graphics chip. No plugin, no app download. It works in every modern browser on desktop and mobile. Libraries like Three.js sit on top of it and make it practical to build with.

That means you can put a genuine 3D object on a web page. Not a video of one. Not a spinning GIF. A live model that renders fresh every frame, reacts to input, and can change colour, material or configuration on command.

For a business, that opens up three things a flat site can't do well:

• Inspection. The buyer rotates, zooms and examines the product as if it were in their hands.
• Configuration. The buyer changes the finish, the colour or the options and sees the result instantly.
• Atmosphere. The brand gets a memorable, living first impression instead of another stock hero image.

Where 3D earns its keep, and where it doesn't

Here's the honest part most agencies skip. 3D is not right for every page. It's a tool with a job, and bolting it onto the wrong page just adds weight.

Where it pays:

• Physical products with detail worth seeing. Watches, furniture, footwear, audio gear, anything where craftsmanship is the selling point.
• Configurable products. Anything with finishes, colours or modules the buyer has to choose between.
• Spaces and architecture. Property, venues, fit-outs and interiors, where understanding the layout matters.
• Launches and brand flagships. Where the goal is to be remembered and talked about.

Where it doesn't:

• Service pages that need to explain and persuade. Words, proof and a clear offer do the heavy lifting there.
• Checkout and forms. Anything between a buyer and the pay button should be light, fast and boring.
• Pages built mainly to rank for search. Search engines read text. A beautiful scene with no content around it is invisible to them.

The rule we use: 3D should answer a buyer's question or create a feeling the brand needs. If it does neither, cut it.

Performance is the catch, and the moat

Most agencies bolt on heavy 3D libraries, tank their load times, and call it "immersive." We do the opposite. The trick is shipping high-fidelity visuals that stay under budget:

• Draco-compressed geometry so models load in a fraction of the bandwidth.
• Lazy-initialized canvases that only spin up the GPU when the element scrolls into view.
• Reduced-motion fallbacks so the experience degrades gracefully for every device and accessibility setting.

Those three are the foundation. On top of them we'd add:

• Compressed textures. Textures are often heavier than the geometry. Right-sized, compressed image formats make a huge difference on mobile.
• A static poster first. The page paints a lightweight image of the scene immediately, then swaps in the live 3D once it's ready. The visitor never stares at a blank box.
• A performance budget. A hard ceiling on file size and frame cost, agreed before a single model is built, and enforced through the project.
• Adaptive quality. Lower resolution rendering and simpler effects on weaker devices, so a mid-range phone gets a smooth experience rather than a stutter.

Why is this a moat? Because it's hard. Anyone can drop a heavy 3D library onto a page. Very few teams can make it load fast, run smoothly on a three-year-old phone and still pass the speed tests Google cares about. If you want to see what's at stake, read why a slow website is a silent business killer and what the Core Web Vitals rules actually measure. A 3D site that fails those is a liability, however pretty it is.

How to build a 3D site that sells, step by step

1. Start with the buyer's question. What does the visitor need to see or feel to say yes? If the answer is "the stitching on the strap," that's your 3D moment. Write it down before anyone opens a design tool.
2. Set the performance budget. Decide the maximum weight and the target devices up front. This prevents the classic "it looked amazing on the designer's machine" disaster.
3. Get or build the models. Many manufacturers already have CAD files. These need to be cleaned, simplified and optimised for the web, because an engineering file is far too heavy to ship.
4. Design the interaction. Drag to rotate, scroll to reveal, tap to change finish. Keep it obvious. If people need instructions, it's too clever.
5. Wrap it in real content. Headline, benefits, proof, price and a clear call to action around the scene. The 3D earns attention. The copy converts it.
6. Build the fallbacks. Poster image, reduced-motion version, and a graceful path for older devices.
7. Test on real phones. Not just the latest flagship. A mid-range Android on a patchy mobile connection is the truth serum.
8. Measure and iterate. Compare the 3D page against the flat version on the numbers that matter, then refine.

From spectacle to sales

A product that rotates in real time, a configurator that previews materials live, a hero that reacts to the visitor, these aren't decorations. They shorten the distance between curiosity and checkout. When a customer can feel the product through the screen, price stops being the conversation.

Here's why that happens. Most hesitation online is uncertainty. Will the colour look right in my room? What does the back look like? Is the finish glossy or matte? Each unanswered question is a reason to leave the tab open and never come back. A live model answers them on the spot, and a buyer with no questions left is a buyer ready to pay.

Configurators go further. The moment someone has chosen their colour, their finish and their size, they've mentally taken ownership. They're no longer browsing a product. They're looking at theirs. That's a very different emotional state to be in when the "Add to cart" button appears.

A worked example: the furniture configurator

Here's a hypothetical. Say you run an Australian furniture maker selling a solid-timber dining table in three timbers, two leg styles and four lengths. Your current product page has a grid of photos, but you can't photograph every combination, so buyers see the oak version and have to imagine the walnut with the tapered legs at 2.4 metres.

They email to ask. Some wait for a reply. Plenty don't.

A WebGL configurator changes the page. The buyer picks walnut, taps tapered legs, slides the length to 2.4 metres, and watches the table update in front of them. They drag it around to see the grain on the end. They zoom into the joinery. The price updates as they choose. When they're done, the "Add to cart" button carries exactly the table they've just built.

What you'd watch after launch: fewer pre-sale emails asking what a combination looks like, more time on the product page, a higher share of visitors reaching the cart, and fewer returns from buyers who got something different to what they pictured. None of those are guaranteed. All of them are measurable, which is the point.

How to measure whether 3D is paying for itself

"It looks incredible" is not a business result. Track these instead:

• Interaction rate. What share of visitors actually touch the 3D element? Low numbers mean it's either hidden, slow or confusing.
• Time on page and scroll depth. Useful, but only alongside conversion. Attention that never converts is just a nicer bounce.
• Add-to-cart and enquiry rate. The real test. Compare against the flat version of the page if you can run both.
• Load and responsiveness metrics. Watch real-user speed data, especially on mobile. If these slip, you're paying for the 3D in lost visitors.
• Return and support rates. Better understanding before purchase should mean fewer "not what I expected" problems after.

And keep the fundamentals in view. The 3D scene is one part of the page. The structure around it still has to follow the rules in the anatomy of a landing page that prints money: one clear promise, proof, and an obvious next step.

The mistakes that turn immersive into invisible

• 3D with no purpose. A floating abstract shape that has nothing to do with the offer. Pretty for three seconds, then ignored.
• Loading the whole scene before the page appears. The visitor sees nothing, assumes the site is broken, and leaves.
• Hijacking the scroll. Scroll-driven scenes that trap people and stop them reaching the content feel clever to designers and infuriating to buyers.
• Ignoring accessibility. No reduced-motion option, no keyboard path, no text alternative. That shuts people out and can make some visitors feel physically unwell.
• Burying the copy. Search engines and answer engines read words. If the only thing on the page is a canvas, you've built something nobody can find.
• Desktop-only testing. Most of your visitors are on phones. Build for them first.

Build the experience, keep the speed

The 2D web was built for documents. Your brand isn't a document. It's an experience, build it like one.

But build it with discipline. Every 3D element should answer a buyer's question or create a feeling your brand needs, load fast on an ordinary phone, and sit inside a page that still sells with words and proof. That combination is rare, which is exactly why it works. Our web development team builds sites that way: immersive where it counts, fast everywhere, and measured on sales rather than applause. If your product deserves to be explored instead of skimmed, let's break the plane.

Frequently asked questions

Does a 3D website slow down page speed?

It can, and many do. It doesn't have to. Compressed geometry and textures, canvases that only start when scrolled into view, a lightweight poster image first, and a strict performance budget keep a 3D site fast. The key is treating speed as a requirement from day one, not something to fix after the models are built. Always test on mid-range phones, not just powerful desktops.

Is WebGL good or bad for SEO?

WebGL itself is neutral. Search engines can't read what's drawn inside a 3D canvas, so the scene contributes little on its own. What matters is the content around it: real headings, descriptive copy, structured data and fast loading. Pair the 3D experience with a properly written page and you get the engagement benefits without sacrificing visibility in search or AI answers.

What kinds of businesses benefit most from a 3D website?

Businesses selling physical products where detail, finish or configuration drives the decision benefit most: furniture, fashion accessories, watches, audio gear, vehicles, and property or interiors. Brands launching something new or wanting a memorable flagship also gain. Service businesses usually get more from sharp copy, proof and a clear offer, with 3D used sparingly for atmosphere rather than as the centrepiece.

Do WebGL websites work on mobile phones?

Yes. WebGL runs in every modern mobile browser. The challenge is performance, because phones have less graphics power and often slower connections. Well-built 3D sites detect the device, lower rendering quality where needed, load compressed assets and fall back to a static image on older hardware. Mobile should be the first design target, since that's where most visitors arrive.

So here is the short version. Positioning comes before the logo because the logo can only express a decision you have already made. Decide who you serve, what you do better than anyone, and what you refuse to do. Then brief the designer. Do it the other way around and you pay for a pretty mark that means nothing.

This guide walks you through exactly how to make that decision, how to test it, and how to turn it into a brief that makes the design work almost obvious.

Positioning is a decision, not a slogan

Your positioning is the space you occupy in your customer's mind relative to every alternative. It's a deliberate trade-off. Trying to be for everyone is the fastest way to be memorable to no one.

Notice the word "alternative". Your customer is never judging you in a vacuum. They are comparing you to the competitor down the road, the cheaper option online, the DIY version, and the very real option of doing nothing at all. Positioning is the answer to one question they ask, often without realising it: why this one instead of the others?

A slogan is a line of copy. A position is a commitment that shows up in your pricing, your offer, who you hire, what you say no to and how you answer the phone. If it lives only on the homepage banner, it is not a position. It is decoration.

If your brand could be swapped for a competitor's with a logo change, you don't have a brand. You have a vendor.

Why founders skip it and go straight to design

Because design feels like progress. You can see a logo. You can put it on a shirt, show it to your partner, post it on Instagram. Positioning is invisible, uncomfortable and full of decisions that feel like losing customers.

There is a second reason. Most founders believe their position is already obvious. "We do quality work at a fair price with great service." Every single competitor says exactly the same thing. That is not a position. That is the entry ticket.

The cost of skipping it shows up later, and it is expensive. You pay a designer for three rounds of concepts, hate them all, and cannot say why. The website copy goes through twelve drafts because nobody agrees what the business is. The ads sound like everyone else's ads. You end up competing on price because price is the only difference a customer can see. We unpack that last trap in why "being the best" is quietly killing your business, and it almost always starts here.

The four questions that build a position

• Who is this truly for? Name the person. Specificity is magnetic.
• What do they believe that the market ignores? Tension creates attention.
• What is the one thing you do better than anyone? Pick the hill.
• What are you willing to be bad at? The bravest brands say no out loud.

Each one deserves a proper answer, so let's go through them.

Who is this truly for?

Not "small businesses". Not "women 25 to 45". A person. What do they do all day, what keeps them up at night, what have they already tried that failed? The tighter the description, the easier every later decision gets. You are not excluding everyone else from buying. You are deciding who you are built around.

What do they believe that the market ignores?

Every category has an unspoken frustration. Accountants who only talk to you at tax time. Gyms that sell memberships to people they know will quit. Builders who never return calls. When you name the thing your customer already feels and nobody else says, you stop sounding like a vendor and start sounding like someone who gets it.

What is the one thing you do better than anyone?

One thing. Not a list of seven strengths. Speed, depth in a single niche, a guarantee nobody else will match, a method you invented. If you cannot pick one, pick the one your best customers mention first when they recommend you. That is usually the truth.

What are you willing to be bad at?

This is the question founders dodge, and it is the one that makes the rest believable. If you are the fastest, you are probably not the cheapest. If you are the most bespoke, you are probably not the quickest. Saying the trade-off out loud tells customers you are serious and stops you drifting back to the middle every time a wrong-fit prospect asks for a discount.

How to find your position in a week

You do not need a six-month brand strategy engagement to get most of the way there. You need honest inputs and the discipline to choose. Here is a process any founder can run.

1. Talk to five of your best customers. Ask why they chose you, what they almost chose instead, and how they would describe you to a friend. Write down their exact words. Their language beats yours every time.
2. Map the alternatives. List every option your customer genuinely considers, including doing nothing and doing it themselves. Next to each one, write what it is known for.
3. Find the gap. Look for the combination nobody owns. Maybe everyone is either cheap and slow or premium and slow. Premium and fast might be sitting there empty.
4. Draft the positioning statement. One sentence: for [who], who [feel this tension], we are the [category] that [one thing you do best], unlike [main alternative]. It will be clunky. That is fine. It is an internal tool, not a tagline.
5. Write the "not for" list. Three types of customer you will no longer chase. If this feels painful, you are doing it right.
6. Pressure test it. Read the statement aloud to someone who knows your market. If they say "so what?" or "everyone says that", go back to step three.

Most of the value is in steps one and five. Customer words give you the truth. The "not for" list gives you the courage.

A worked example: the bakery that stopped competing on sourdough

Say you run a small bakery in Brisbane. The first instinct is to call it "artisan", hire a designer, and get a hand-drawn wheat logo in muted earthy tones. So does every other bakery within ten kilometres. Your customer now sees four near identical logos and chooses the closest one.

Now run the four questions. Who is it truly for? After talking to regulars, you notice your best customers are office managers ordering catering for team meetings, not weekend sourdough buyers. What do they believe the market ignores? That ordering catering is a stressful chore, with late deliveries, wrong orders and invoices that confuse the accounts team. What do you do better than anyone? You deliver on time, every time, with one simple invoice. What are you willing to be bad at? You will not be the trendiest weekend brunch spot.

Suddenly the design brief writes itself. The brand should feel dependable, crisp and professional, not rustic. The website should lead with an online order form and a delivery guarantee, not a moody photo of flour. The name, the colours and the photography all point at one promise: the catering order that never goes wrong. That is a position a competitor cannot copy by changing their logo, because it lives in how the business actually runs.

This is a hypothetical, but the pattern is real. The position did not come from the design. The design came from the position.

Then, and only then, design

Once the position is locked, every creative decision becomes obvious. The logo, the voice, the photography, the pricing, they stop being matters of taste and become expressions of strategy. That's how you build a brand people don't just recognise, but choose.

Here is how the position translates into each creative decision.

• Name. If you are still choosing one, the position tells you which feelings the name must carry and which it must avoid. Our guide on how to name a brand people actually remember goes deeper on this.
• Logo and colour. A position built on reliability wants restraint and clarity. A position built on rebellion wants something that breaks the category's visual habits. Look at what competitors use and decide whether you are joining or deliberately standing apart.
• Voice. Formal or casual, technical or plain, playful or blunt. The customer you named in question one decides this, not the founder's personal taste.
• Photography and video. Show the proof of your one thing. If it is speed, show it. If it is craft, get close enough to see it.
• Pricing. Your price is part of your position. Premium positioning with bargain pricing confuses people. Budget positioning with premium pricing insults them.

When a designer receives a brief like this, the concept rounds shrink. Feedback stops being "I don't love the blue" and becomes "does this feel dependable to an office manager?" That is a question you can actually answer.

The mistakes that quietly undo a good position

• Positioning on a feature anyone can add. "We offer free delivery" lasts until a competitor offers free delivery. Anchor on something structural, like a method, a niche or a promise that is costly for others to copy.
• Hedging the language. "Mostly for" and "among other things" water everything down. Commit in public, even if you still take the occasional off-profile job in private.
• Letting the website disagree. A sharp position on the homepage with a services page listing twenty unrelated offers tells the customer you did not mean it.
• Changing it every quarter. Positions take time to settle into a market's memory. Refine the wording, keep the core.
• Confusing a message problem with a positioning problem. Sometimes the position is right and the words are wrong. We covered that distinction in your competitors aren't the problem, your message is.

How to know your positioning is working

You will not see it in a single metric, but the signs are consistent and easy to spot.

• Prospects repeat your one thing back to you on the first call, before you have said it.
• More enquiries come from the customer you named, and fewer from the ones on your "not for" list.
• You get asked for discounts less, because the comparison is no longer purely on price.
• Referrals describe you the same way you describe yourself.
• Your team makes small decisions without asking you, because the position tells them what the brand would do.

If none of that is happening after a few months, the position is either too vague or not showing up in the actual experience. Both are fixable, and both are cheaper to fix than a rebrand.

What to do before you brief a designer

Do the interviews. Write the statement. Write the "not for" list. Put all three on one page and hand that page to whoever is designing your brand. If they do not ask for something like it, that tells you something about the designer too.

This is exactly the work we do with founders at LIQID NOISE before a single colour gets picked. Our brand and business consulting starts with the hard questions, then carries the answers through to the name, the identity, the website and the ads, so every piece says the same thing. A brand that knows what it stands for is cheaper to market, easier to design, and much harder to copy.

Get the position right and the logo becomes the easy part.

Frequently asked questions

What is the difference between brand positioning and branding?

Positioning is the strategic decision about who you serve, what you do better than any alternative, and what you give up to do it. Branding is how that decision gets expressed: the name, logo, colours, voice, photography and customer experience. Positioning is the brief. Branding is the response. Without a clear position, branding choices come down to personal taste, which rarely persuades a customer.

Should a small business do positioning before getting a logo?

Yes. A logo designed without a position can only look nice, it cannot say anything specific. Even a week of customer interviews and a one-sentence positioning statement gives a designer something real to work with. It also cuts revision rounds, because feedback can be judged against the position instead of against preferences. The logo is cheaper and better when it comes second.

How do you write a positioning statement?

Use a simple internal template: for [a specific customer], who [feels a specific frustration], we are the [category] that [the one thing you do best], unlike [the main alternative]. Keep it to one sentence and fill it with your customers' own words from interviews. It is a decision tool for your team, not a public tagline, so clarity matters more than polish.

How often should you change your brand positioning?

Rarely. A position takes time to settle into your market's memory, and changing it often resets that work. Refine the wording as you learn, and revisit the whole thing only when something fundamental shifts, like a new core customer, a new offer or a major change in competitors. Constant repositioning usually signals the original decision was never really made.

Except a website is not a cost. It is the front door to your revenue. And a cheap front door does not save you money, it silently leaks it, month after month, in amounts that dwarf whatever you "saved" up front.

So what does a cheap website really cost? The real cost isn't the build price. It's the customers the site fails to convert, every month, for as long as it stays live. A site that turns fewer visitors into enquiries costs you far more in lost revenue than a better build would ever have cost to make. Here's how to work out your own number, what cheap actually gets you, and what to demand instead.

Do the math nobody shows you

Say your site gets 1,000 visitors a month. A cheap, generic, slow website might convert 1% of them into a lead. A sharp, fast, conversion-engineered site converts 3%. That is not "a bit better." That is triple the customers from the exact same traffic.

You did not save money on a cheap website. You just agreed to pay for it, quietly, forever, in lost customers.

Now multiply those missed leads by your average customer value, then by twelve months, then by the years that site stays up. The "expensive" website was never the one that cost you the most. It was the cheap one.

Let's put numbers on it. Stick with the same hypothetical 1,000 visitors a month. At 1%, you get 10 leads. At 3%, you get 30. Say one in five leads becomes a customer, and the average customer is worth $2,000 to you. The cheap site produces 2 customers a month, $4,000. The sharp site produces 6 customers a month, $12,000. That's a gap of $8,000 a month, or $96,000 a year, from the same traffic and the same ad spend.

Your numbers will be different. Maybe you get more traffic, maybe your customers are worth less. Run the sum anyway. Almost every business that does it finds the difference between a site that sells and a site that just exists is worth many times the cost of building it properly.

What cheap actually buys you

• A template a thousand other businesses are using, so you look like everyone else and command nobody's premium.
• Bloated, slow-loading code that bleeds visitors before your headline even appears.
• No conversion strategy, just pages that exist, with no plan to turn a reader into a buyer.
• A mobile experience that fights the user, where most of your traffic actually lives.
• Zero differentiation, which means you are forced to compete on price. The most expensive place to compete.

Cheap web design does not just underperform. It actively trains your market to see you as a cheap option. And once you are the cheap option, every other part of your business gets harder.

The hidden costs that show up later

The lost conversions are the big one. But cheap builds carry a list of other costs that arrive months after the invoice is paid.

• Speed penalties. Slow sites frustrate visitors and can hold back your search rankings, since Google says its core ranking systems look to reward a good page experience. We cover the details in a slow website is a silent business killer.
• Wasted ad spend. Every dollar of paid traffic sent to a page that doesn't convert is a dollar you paid twice for: once for the click, once for the lost sale.
• Plugin debt. Cheap builds often stack plugins to add features. Each one needs updates, can conflict with the others and adds weight to the page.
• Security holes. Out-of-date themes and plugins are a common way sites get hacked. Cleaning up after a breach costs far more than preventing one.
• Lock-in. Some budget builders own your hosting, your domain or the site itself. Leaving can mean starting from scratch.
• Your time. Every hour you spend fighting a clunky editor or chasing a developer who's stopped replying is an hour not spent on the business.
• The rebuild. Most cheap sites get replaced within a few years. You pay for the website twice, and lose the time in between.

Why cheap sites convert worse

It's not magic. A high-converting site is the result of dozens of deliberate decisions that a rushed template job skips. The difference usually comes down to these:

1. A clear promise above the fold. Within seconds, a visitor knows what you do, who it's for and why you're the right choice. Template sites lead with a stock photo and a vague slogan.
2. A single, obvious next step. One primary call to action, repeated at the right moments. Not five competing buttons.
3. Proof where doubts appear. Reviews, results, logos and guarantees placed right next to the moments a visitor hesitates.
4. Copy that sells. Words written around your customer's problems and objections, not a list of your services.
5. Speed on real phones. Lean code and optimised images, tested on mobile connections, not just a fast office laptop.
6. Frictionless forms. Only the fields you really need, clear labels and a fast confirmation.
7. Measurement. Analytics and conversion tracking set up from day one, so you can see what's working and improve it.

Your website should work like your best salesperson, not an unattended brochure rack. That's the whole idea behind your website is your best salesperson.

Premium is a decision, not a splurge

The brands that dominate their markets do not think of their website as an expense to minimise. They think of it as a machine to optimise, because they know the site that converts 3% instead of 1% will out-earn its own cost many times over before the year is out.

That doesn't mean spending as much as possible. An expensive site can be just as unprofitable as a cheap one if it's all animation and no strategy. The goal is return. You want every dollar spent on the build to come back as more leads, higher-value customers or lower acquisition costs. Price is what you pay. Profit is what you should be judging.

It also means the launch isn't the finish line. A profitable site gets better after it goes live. You watch where visitors drop off, test a sharper headline, move the reviews higher, cut a form field, speed up a heavy page. Small, measured improvements compound. A cheap site is usually built once and abandoned, so it gets worse every year as your business, your customers and the web move on without it.

How to tell if your current site is costing you

You don't need a developer to spot the warning signs. Spend thirty minutes on this checklist:

1. Open your site on your phone, on mobile data, not wifi. Count how long it takes before you can actually read and tap things.
2. Run it through Google PageSpeed Insights. It's free and shows how your site performs on real-world speed measures. Our guide to Core Web Vitals in 2026 explains what the scores mean.
3. Do the five-second test. Show your homepage to someone who doesn't know your business for five seconds, then ask what you do and who it's for.
4. Try to become a customer. Fill in your own form, click your own buttons, book your own appointment. Note every moment of friction.
5. Check your analytics. Look at your conversion rate from visitor to enquiry. If you don't know it, that's your first problem.
6. Compare against your top competitor. Put the two sites side by side. Which one would you trust with your money?

A worked example: the plumber's rebuild

Say you run a plumbing business in Newcastle. Your site cost a few hundred dollars off a template years ago. It gets steady traffic from Google and a small monthly ad budget, but the phone doesn't ring much from it.

You run the checklist. On mobile, the homepage takes ages to load and opens on a stock photo of a tap. The phone number is tiny and buried in the footer. The contact form asks for nine fields. There are no reviews anywhere, even though you have dozens of good ones on Google.

A proper rebuild fixes each of those. A fast, lean site. A headline that says "Emergency plumber in Newcastle, on site fast." A tap-to-call button that follows the visitor down the page. A three-field quote form. Real reviews and photos of real jobs next to every call to action. Same traffic, same ad spend. But now the site actually closes the visitors it already had. That's the entire argument for spending properly: you stop leaking the customers you're already paying to attract.

What to ask before you hire anyone

• How will this site generate more leads or sales, specifically?
• Will I own the domain, the hosting account and the site files?
• How fast will it load on mobile, and how will you measure that?
• Who writes the copy, and is it written to convert?
• What tracking will be set up so we can measure results?
• What happens after launch? Who handles updates, security and improvements?

If the answers are vague, the site will be too. And if you want the page-level playbook, the anatomy of a landing page that prints money shows exactly what a converting page is made of.

Build a front door worth walking through

We do not build cheap websites, and we do not build expensive ones. We build profitable ones, sites engineered to return more than they cost, on purpose, measurably.

Our web development team designs and builds fast, conversion-focused sites for businesses that want their website to pull its weight. If you are tired of quietly paying the cheap-website tax, let's build you a front door worth walking through.

Frequently asked questions

How much should a small business website cost?

It depends on what the site needs to do. A simple brochure site costs less than one with bookings, e-commerce or custom features. Instead of starting with price, start with value: estimate how many extra leads or sales a better site could produce and what they're worth. Spend in proportion to that, and always ask how the build will make you money.

Are website templates bad for business?

Not automatically. A well-chosen template, customised properly with strong copy, fast hosting and a clear conversion path, can work well for a new business. The problem is templates used as a shortcut: generic design, bloated code, stock images and no strategy. That combination looks like everyone else and converts poorly. The template isn't the issue. The lack of thinking is.

How do I know if my website is losing me customers?

Check your conversion rate from visitor to enquiry or sale, test your site speed on a phone, and try to complete your own contact form or checkout. High traffic with few enquiries, slow mobile load times, confusing navigation and missing reviews are all warning signs. If you can't measure conversions at all, fixing your tracking is the first step.

Is it worth rebuilding a website that still works?

"Works" isn't the standard. A site can load and still quietly lose you customers every day. If it's slow, hard to use on mobile, lacks proof and has a weak call to action, a rebuild often pays for itself. Estimate the extra leads a better site would bring, multiply by customer value, and compare that to the rebuild cost.

Here is the short answer. A brand video that sells follows a deliberate story structure: a hook that earns attention, stakes the viewer recognises, a transformation where your brand is the bridge, and a resolution that lands a feeling and points to one next step. The camera, the lighting and the grade only multiply whatever that structure already delivers. Get the story wrong and no amount of gear will save it.

That is uncomfortable news if you just spent your budget on a drone and a gimbal. It is great news if you are willing to do the thinking first, because thinking is cheap and almost nobody does it.

Story is the engine, visuals are the chassis

Stunning shots without structure are a screensaver. The videos that move people, and move product, follow a deliberate emotional arc.

Think about what actually happens when someone watches your video. They are not assessing your lens choice. They are asking, half consciously, three questions. Is this about me? Should I care? What do I do now? A screensaver answers none of them. It is pleasant, it is expensive, and it is forgotten the second the next post loads.

Most brand videos fail the same way. The brief says "we want something cinematic that shows who we are." So the crew shoots slow-motion hands, a sunrise over the office, the founder walking down a corridor, a logo at the end. Every frame is gorgeous. Nothing happens. There is no tension, so there is no reason to keep watching, and there is no change, so there is nothing to remember.

Nobody shares a video because the colour grade was nice. They share it because it made them feel something.

The structure that converts

• The hook: Earn the next five seconds with tension, beauty, or a question.
• The stakes: Show the world your audience lives in, the problem, the desire, the gap.
• The transformation: Position your brand as the bridge, not the hero. The customer is the hero.
• The resolution: Land the feeling, then point to the next step.

Each beat has one job. Here is how to make each one pull its weight.

The hook

You are not competing with other brand videos. You are competing with everything else in the feed, so the first frame has to create a gap the viewer wants closed. An unexpected image. A line of dialogue that starts mid-argument. A question the viewer has asked themselves. Never open on your logo. A logo answers a question nobody asked. If you want the thinking behind that first frame, read our breakdown of why some images stop the scroll and others vanish. The same psychology runs the opening seconds of a video.

The stakes

This is where most brand films are thinnest, because it means showing a problem, and marketing teams hate showing problems. But without stakes there is no story. Show the frustration, the wasted Saturday, the moment of doubt, the thing your customer wants and cannot quite reach. The viewer should recognise their own life before they recognise your product.

The transformation

Your brand enters as the guide, the tool, the bridge. Not the star. The moment you make the video about how great your company is, the viewer switches off, because now it is an ad about you rather than a story about them. Show the change happening in the customer's world, and let your product be the reason it became possible.

The resolution

Land the feeling first. Relief, pride, excitement, calm. Hold on it long enough to register. Then, and only then, give one clear next step. One. A brand video that ends with a website, a phone number, three social handles and a discount code is asking the viewer to make four decisions at the exact moment they are ready to make one.

Craft amplifies the message

Once the story is right, production value multiplies its impact. Intentional lighting, motivated camera movement, sound design, and pacing aren't decoration, they're how you control emotion frame by frame. We start every shoot with the story, then bring the cinematic craft to make it unforgettable.

Here is what that looks like in practice.

• Lighting sets the mood before anyone speaks. Hard, contrasty light reads as tension. Soft, warm light reads as safety. Use the stakes section and the resolution to shift from one to the other and the audience feels the change without being told.
• Camera movement needs a motive. A push in says "pay attention, this matters." A handheld shot says "this is real and happening now." Movement for its own sake just says "we hired a gimbal."
• Sound is half the picture. Room tone, the click of a lid, a breath before a line. Viewers forgive soft focus far faster than bad audio. Music should be chosen for the arc, not for the vibe of the brand deck, which is why we argue that generic royalty-free music quietly cheapens a film.
• Pacing is the edit's job. Tight in the hook, slower in the stakes so they land, a lift into the transformation, space in the resolution. Pacing that never changes flattens every emotion into the same beat.

How to brief a brand video that works

The brief decides the film. A vague brief produces a vague video, however talented the crew. Before you talk to anyone about cameras, answer these on one page.

1. Who is the one viewer? Not "our audience." A specific person, with a specific problem, at a specific moment in their week.
2. What do they believe before watching? Write down the objection or misconception you need to shift.
3. What should they feel at the end? One emotion. If you list five, you have picked none.
4. What is the single next step? Book a call, visit the store, watch the product demo. Pick it now so the story can build toward it.
5. Where will it live? A homepage hero, a paid social ad and a trade show screen need different lengths, openings and aspect ratios. Plan cutdowns before the shoot, not after.
6. What is true? List the real customers, real places and real proof you can put on camera. Truth on screen outperforms polish on screen almost every time.

If you cannot answer these, you are not ready to shoot. That is fine. It just means the first dollar should go into strategy, not production.

A worked example: the Brisbane bike shop

Say you run a hypothetical independent bike shop in Brisbane and want a 60-second brand film.

The screensaver version: golden-hour riders on a trail, close-ups of chrome, the shop sign, the logo. Pretty. Forgettable.

The story version starts with a hook: a rider on the side of the road, chain snapped, a ride group disappearing over the hill. Five seconds and you already want to know what happens. The stakes: this person almost quit riding because their last bike never fitted and never got fixed properly. The transformation: a fitting session in your shop, a mechanic who explains what went wrong, the bike adjusted to the rider rather than the other way round. The resolution: the same rider back in the group, at the front this time, grinning. End card, one line, one action: "Book a free bike fit."

Same location. Same crew. Same budget. One version decorates the shop. The other sells the one service that brings in new customers. The only difference is the thinking done before anyone pressed record.

Common mistakes that kill brand videos

• Making it about the company. Office tours and founder monologues about "our passion" rarely convert. Make the customer the hero.
• Committee scripting. Every stakeholder adds a message until the film says nothing. Protect the single idea.
• Too long for the placement. A three-minute film may work on a landing page and die in a feed. Cut for where it will actually be watched.
• No sound plan. Many people watch on mute. Burned-in captions and visual storytelling keep the story alive without audio.
• Shooting once, using once. A proper shoot should feed months of content. We lay out how in one shoot, 90 days of content.

How to measure whether it worked

"People loved it" is not a result. Decide what success means before launch, based on where the video lives.

• In paid social: watch the drop-off in the first few seconds to judge the hook, then hold rate through the middle to judge the stakes, then clicks or conversions to judge the resolution.
• On a landing page: compare conversion rate with the video against the page without it. If the video does not lift the page, it is decoration.
• On the homepage: watch time and what visitors do next. A good brand film pushes people deeper into the site, not back to Google.

Each metric maps to a beat of the story. When a number is weak, you know exactly which part of the film to fix instead of guessing. A weak hook is a reshoot of five seconds. A weak ending is a new end card. Neither needs a whole new film.

Story first, then make it beautiful

Gear is a commodity now. Anyone can rent a cinema camera for a weekend. What cannot be rented is a clear point of view about who your customer is and what change you create for them. That is the part that sells, and it is the part most brands skip.

Our photo and video team works in that order every time: story, structure, then cinematic craft. If you want a brand film that people finish, remember and act on, start with the thinking, then make it beautiful.

Frequently asked questions

How long should a brand video be?

As long as the story needs and no longer, cut to the placement. A homepage or landing page film can run longer because the viewer chose to be there. Feed placements need a tight hook and a shorter edit. The smart move is to plan one core film plus shorter cutdowns during pre-production, so every version keeps a complete arc instead of being chopped at random.

What is the difference between a brand video and an ad?

An ad usually pushes one offer toward an immediate action. A brand video builds belief: who you help, what changes for them and why you are the right bridge. The best brand videos still end with a clear next step, so the line is blurry. Both need a hook, stakes and a resolution. The brand film simply spends more time on the feeling.

Do I need a cinema camera to make a cinematic brand video?

No. Cinematic comes from intention, not equipment. Motivated lighting, deliberate composition, clean sound and pacing that follows the story matter far more than the sensor. A modern phone with good light and good audio can produce a film that feels crafted. An expensive camera pointed at a weak story still produces a screensaver.

Should the founder appear in the brand video?

Only if they serve the customer's story. A founder explaining why they built the solution to a problem the viewer recognises can be powerful. A founder talking about the company's values for two minutes usually is not. Keep the customer as the hero and use the founder as the guide who makes the transformation believable.

So what does SEO that survives AI search actually look like? It's content built to be quoted, not just ranked. Clear entities, direct answers, real expertise and structured data, organised into deep topic clusters. Do that and you show up in Google's results and inside the answers ChatGPT, Perplexity and AI Overviews hand your buyers.

From ranking to being cited

The new goal isn't position one. It's being the source the AI quotes. That requires a different kind of content, clear, structured, and genuinely authoritative.

Here's the shift in plain terms. Old search gave the user ten options and let them choose. Answer engines make the choice for them. They read a pile of pages, pull out the clearest statements, and stitch them into one confident response. Maybe they show a few sources underneath. Maybe the user clicks one. Often they don't click at all.

That changes what winning means. A page that ranks fourth but contains the single clearest explanation of a topic can be quoted ahead of the page that ranks first. A page that ranks first but buries its point under 400 words of throat-clearing can be skipped entirely. Position still matters, because engines lean on pages they can already find. It just isn't the finish line any more.

If the alphabet soup of AEO, GEO and SEO has your head spinning, we've untangled it in the new alphabet of getting found. The short version: they're one job now, not three.

The four pillars of answer-engine SEO

• Entity clarity: Be unambiguous about who you are, what you do, and how you connect to related topics. Schema markup is no longer optional.
• Direct answers up top: Lead with the conclusion, then support it. Machines extract the first clear statement that answers the query.
• Genuine expertise: First-hand experience, real data, and named authors signal trust that thin AI-spun content can't fake.
• Structured data everywhere: FAQs, how-tos, products, and reviews marked up properly become machine-readable assets.

Each of those pillars answers a question the machine is silently asking. Who is this? What's the answer? Why should I believe it? Can I read it without guessing? Miss one and you give the engine a reason to quote someone else.

Write for the human, structure for the machine. The brands that do both will own the next era of search.

Pillar by pillar: what to actually change

Principles are cheap. Here's what each pillar looks like when you sit down and edit your site.

Entity clarity. Your business name, what you do, where you do it and who you do it for should read identically on your homepage, your about page, your Google Business Profile, your social bios and every directory listing. Add Organization schema to your site and link it to your real profiles. When an engine sees five sources agreeing, it gets confident. When it sees five slightly different stories, it hedges or moves on. We go deeper on this in how to make AI trust your brand as an entity.

Direct answers up top. Take your ten most important pages and read the first paragraph of each. Does it answer the question the page exists for? Most don't. They warm up, set the scene, tell you how passionate the company is. Rewrite so the first two sentences give the answer plainly, then spend the rest of the page proving it. Use question-shaped headings that match how people actually ask.

Genuine expertise. Put a real name on your content, with a short bio that says why that person knows the topic. Include things only someone who does the work would know: the mistake clients always make, the step that takes longer than expected, the reason one approach beats another in practice. That's the texture generic content can't imitate, and it's exactly what makes a passage worth quoting.

Structured data everywhere. Mark up what you have. FAQ sections, how-to steps, products with price and availability, reviews, your business details, articles with authors and dates. Structured data doesn't guarantee a citation. It removes the guesswork, so the engine spends its confidence on your content rather than on decoding your layout.

Topical authority beats one-hit content

Answer engines reward depth. A single great article is a fluke, an interconnected cluster covering every facet of a topic signals that you own the subject. We build content architectures, pillar pages, supporting articles, and internal links, so your domain reads as the definitive voice in your niche.

Here's how a cluster is built, in order:

1. Pick one topic you want to be known for. Not five. One commercially important subject your buyers research before they buy.
2. List every question a buyer asks about it. Pull them from sales calls, enquiry emails, your team's heads and the "people also ask" boxes in search. Aim for twenty or more.
3. Write the pillar page. One comprehensive guide that covers the topic end to end and answers the big questions directly.
4. Write supporting articles. Each one takes a narrower question and answers it properly, with its own direct answer up top.
5. Link them deliberately. Every supporting article links to the pillar and to its closest siblings, using natural, descriptive anchor text. The pillar links out to all of them.
6. Keep it current. Review the cluster every few months. Update facts, prices and examples, and add new questions as buyers start asking them.

The result is a site where any question on that topic lands on a page that answers it, and every page vouches for the others. That pattern is what engines read as authority.

A worked example: the accountant who stopped chasing keywords

Say you run a small accounting firm in Perth that works with tradies. For years your SEO plan was a list of keywords like "accountant Perth" and "tax return tradie," and a blog post every month targeting one of them. Some ranked. None were ever quoted by an AI assistant, because none of them clearly answered anything.

Now you switch approach. You pick one topic: tax for sole trader tradies. You list the questions your clients actually ask. What can I claim for my ute? Do I need to register for GST? How much should I put aside for tax each week? What happens if I miss a BAS deadline?

You write a pillar guide covering the whole subject, then a supporting article for each question. Every article opens with a two-sentence plain answer. Each one is written and signed by the partner who handles those clients, with a line about her experience. Each has FAQ schema. They all link to each other.

When a tradie asks an assistant "what can a sole trader electrician claim on tax," the engine now finds a page on your site that answers exactly that, written by a named expert, backed by a cluster of related pages. That's a page it can quote with confidence. You haven't stopped doing SEO. You've done it in a way the new engines can use.

Notice what didn't happen in that example. No new website. No huge budget. No trick. The firm took knowledge it already had in the heads of its team and published it in a shape machines can read. Most businesses are sitting on the same raw material. The expertise is there. It's just trapped in sales calls and client emails instead of on pages an engine can find.

How to measure SEO when fewer people click

If answer engines resolve more questions without a click, raw organic traffic becomes a weaker scorecard. You need a few extra lenses.

• Prompt checks. Keep a list of the questions your buyers ask. Every month, run them through ChatGPT, Perplexity, Gemini and Google's AI results. Record whether you're mentioned, how you're described and which pages are cited.
• AI referral traffic. In your analytics, look for visits from AI tools as a separate source. It's often small, but it tends to be high intent.
• Branded search. When assistants mention you, people often go and search your name afterwards. Rising branded searches can be a sign the answers are working for you.
• Enquiry quality. Ask new leads how they found you. "ChatGPT recommended you" is a real answer now, and it's worth counting.

The mistakes that keep good businesses invisible

• Burying the answer. Long intros, brand fluff and a point that finally arrives in paragraph six. The engine has already moved on.
• Mass-producing AI content. Hundreds of thin, generic pages don't signal authority. They signal that nobody on your team knows anything the model doesn't.
• Inconsistent business details. Different addresses, service areas or descriptions across the web make you a risky source to cite.
• Orphan articles. Great posts with no links to or from related content look like one-offs, not expertise.
• Blocking the crawlers you want. Check your robots file and firewall settings. Some sites quietly block the AI crawlers they're hoping will cite them.

Search is changing, the fundamentals are not

Search is changing faster than it has in twenty years. The fundamentals, clarity, authority, and structure, have never mattered more.

That's the good news hiding in all this disruption. The businesses that win in AI search aren't the ones with secret tricks. They're the ones who say clearly what they do, prove they know it and make their pages easy for a machine to read. If you want the detail on writing passages engines love to lift, read how to write content AI can't help but quote.

If you'd rather have a team do it, that's the core of our digital marketing work: entity cleanup, content architecture, structured data and the ongoing prompt checks that tell you whether it's working. Start with one topic, build it properly and watch where you start showing up.

Frequently asked questions

Is SEO still worth it with AI search?

Yes. AI answer engines still pull from pages they can find, crawl and trust, so the basics of SEO feed directly into AI visibility. What changes is the target. Instead of only chasing rankings, you write content with clear direct answers, strong expertise signals and structured data, so it can be quoted inside AI answers as well as ranked in traditional results.

How do I get my website cited by ChatGPT or Perplexity?

Make your pages easy to quote and easy to trust. Answer the main question in the first two sentences, use question-shaped headings, add schema markup, name a real author with relevant experience and keep your business details consistent across the web. Then build clusters of related articles that link to each other, so the engine sees depth on the topic rather than one isolated page.

Does schema markup help with AI Overviews?

Schema markup helps machines understand what a page contains, who wrote it and what business it belongs to. It won't guarantee an AI Overview citation, but it removes ambiguity, which makes your content easier to extract and verify. FAQ, how-to, product, review, article and organisation markup are the most useful places to start for most small businesses.

What is topical authority in SEO?

Topical authority is the signal that your site covers a subject thoroughly, not just once. You build it with a pillar page that covers the topic end to end, supporting articles that each answer a specific question, and deliberate internal links between them. Search engines and AI systems both treat that depth as evidence you're a reliable source on the subject.

Here's the straight answer. Traditional publishing suits authors who want an advance, bookstore reach and an imprint's credibility, and who can accept less control and a long wait. Self-publishing suits authors who want speed, ownership and a bigger share of every sale, and who are willing to fund and run the work themselves. Neither is the "real" way anymore. The right path depends on what the book is for.

And that's the question most authors never ask. They pick a path based on ego or fear, then spend years fighting the wrong system. Let's fix that before you write another chapter.

Start with what the book is for

Before you compare contracts and platforms, get brutally honest about the job this book has to do. Because a book written to win you clients plays by completely different rules from a novel you hope sits in every airport.

• Is it a business asset? A book that positions you as the authority, opens doors to speaking, and gets handed to prospects. Here, speed and control usually beat prestige. We made the case in your book is the most powerful business card you'll ever print.
• Is it a mass-market bid? Fiction, memoir or big-idea nonfiction you want in physical bookstores, reviewed widely and stocked in libraries. This is where a traditional deal earns its keep.
• Is it a legacy project? Something for family, a community or a niche readership. Self-publishing gets it done well without waiting for anyone's permission.

Write the purpose down in one sentence. Every decision after this gets easier.

Traditional publishing: prestige and reach

A traditional deal brings an advance, editorial muscle, bookstore distribution, and the credibility of a recognised imprint. The trade-offs are real: you surrender creative control, accept a small royalty percentage, and wait, often years, on someone else's timeline.

Here's how the path usually works, so you know what you're signing up for:

1. Finish and polish. For fiction, you generally need a complete, polished manuscript. For nonfiction, a strong book proposal with sample chapters and a clear market case is often the entry ticket.
2. Find a literary agent. Most major publishers don't accept unsolicited submissions. You query agents, who pitch your book to editors.
3. Get an offer. If an editor buys it, you negotiate an advance, royalties, rights and delivery dates.
4. Edit and produce. Rounds of editing, cover design, typesetting and printing, on the publisher's schedule.
5. Publish and promote. The publisher handles distribution. Marketing support varies a lot, and many authors are still expected to drive much of their own promotion.

That last point surprises people. A deal does not guarantee a marketing campaign. The publisher bets on the book, but the author's platform often decides how hard they push it.

The other thing to understand is the advance. It's an advance against royalties, meaning you don't earn more until the book's royalties pay back that amount. Plenty of books never do. That's not a failure, the advance is yours to keep, but it means the upfront cheque may be most of what you ever see.

Traditional publishing rents you reach. Self-publishing makes you the owner.

Self-publishing: control and economics

• You keep the rights and the lion's share of every sale.
• You set the schedule, publish in months, not years.
• You own the audience and the data, which compounds across future books.
• You carry the burden too, editing, design, and marketing are yours to fund and manage.

That last bullet is where self-published books live or die. When you self-publish, you become the publisher. That means you're hiring, briefing and paying for the same work a publishing house would do. Skip it and readers can tell within a page.

The production checklist looks like this:

1. Developmental edit. The big-picture pass on structure, argument and flow.
2. Copyedit and proofread. Sentence-level clarity, consistency, spelling and typos.
3. Cover design. Your number one sales tool. It has to look at home next to the best books in your genre.
4. Interior layout. Clean typesetting for print and properly formatted ebook files.
5. ISBNs and metadata. Titles, subtitles, categories, keywords and descriptions that help readers find you.
6. Distribution. Print-on-demand and ebook platforms put your book on the major online stores without a warehouse full of stock.
7. Launch and marketing. Reviews, email, ads, events and media. This never stops at launch day.

Owning the audience is the part most authors underrate. When a publisher sells your book through a retailer, you rarely learn who bought it. When you sell direct, or send readers to a free bonus chapter that asks for an email, every reader becomes a person you can reach again. That list is what makes book two, a course, a workshop or a speaking tour far easier to sell. The book stops being a one-off product and becomes the front door to everything else you do.

Done right, that's a real investment. But you own the result forever, and every future book builds on the audience you've gathered.

The real trade-offs, side by side

Strip away the mythology and the decision comes down to five things.

• Money up front vs money over time. Traditional pays an advance and a smaller cut per copy. Self-publishing costs you money up front and pays a bigger cut per copy.
• Control. Traditional publishers typically have final say on title, cover and timing. Self-published authors decide everything.
• Speed. Traditional timelines commonly stretch past a year after the deal. Self-publishing can move as fast as your production team.
• Reach. Physical bookstore placement is far easier with a traditional publisher. Online, the playing field is much flatter.
• Credibility. A known imprint still carries weight with some media and events. But a well-produced self-published book with real reviews reads as professional to almost every reader.

Notice what's missing from that list: quality. Both paths can produce excellent books. Both can produce bad ones. The path doesn't decide quality. The work does.

The hybrid reality

For most authors today, the smart path is treating your book like a business: professional editing, a cover that competes on the shelf, and a launch strategy built before you publish. Whether you chase a deal or go independent, the work that sells books is the same.

More and more authors now move between both worlds. Some self-publish first, build an audience and real sales, and then take that proof to agents. Others sell print rights to a publisher and keep ebook or audio rights. Some business authors self-publish their signature book and license translations later.

One warning here. "Hybrid" is also a label used by some companies that charge authors to publish. A few are legitimate, offering real editorial and distribution services for a fee. Others are vanity presses selling hope. Before you pay anyone, check exactly what you get, who owns the rights, what royalties you keep, and whether you can see books they've produced that genuinely hold up on the shelf. In traditional publishing, money flows to the author, not from them.

A worked example: the consultant's book

Say you're a Brisbane business consultant. You want a book that explains your method, gets you on stages and warms up prospects before a sales call.

Option one: you write a proposal and query agents. Maybe you land a deal, maybe you don't. If you do, the book might come out well over a year later, with a title and cover you didn't fully pick. Meanwhile, you can't hand it to anyone.

Option two: you self-publish with a professional team. You spend a few months writing with an editor, invest in a strong cover and clean layout, and launch to your existing email list and network. You buy author copies at print cost, send them to your best prospects and hand them out after every talk. The book starts doing its job within months, not years.

For this author, self-publishing wins easily. Now flip it. Say you're a novelist who dreams of your book in every bookstore front window. The traditional route, patience and all, may be the only realistic path to that specific goal. Same decision framework, opposite answer. If you're a time-poor founder in the first camp, the 90-day book shows how to actually get it written.

Mistakes that sink books on either path

• Skipping the edit. Your friends saying it's great is not an edit.
• A homemade cover. Readers judge books by their covers. Every day. On every store.
• No launch plan. Publishing is not launching. Most books die quietly because nobody planned how anyone would hear about them. We covered why in your book launch is where most authors die.
• Signing without reading. Rights, reversion clauses and option clauses matter. Get advice before you sign any contract.
• Waiting for permission. Years spent chasing a deal for a book that was always a business asset is years of lost authority.

What to do this week

1. Write one sentence on what the book must achieve for you.
2. Decide what matters most: advance and bookstores, or speed, control and ownership.
3. Pull ten books in your genre you admire and note who published them and how they look.
4. If you're leaning traditional, research agents who represent books like yours. If you're leaning independent, price out editing, cover and layout.
5. Draft a simple launch plan now, before the manuscript is done.

Whichever door you walk through

The path matters less than the standard. A professional book on either path opens doors. An amateur book on either path closes them quietly.

We help authors produce and launch at a professional standard, whichever door they walk through. Our books and publishing team handles editing, cover design, layout and launch strategy for founders, experts and storytellers who want their book to look like it belongs on the shelf. If you've got a book in you, or half-written on your laptop, let's talk about the right route for it.

Frequently asked questions

Is self-publishing better than traditional publishing?

Neither is better across the board. Self-publishing is better for speed, control, ownership and a larger share of each sale, which suits business books and niche audiences. Traditional publishing is better for advances, physical bookstore distribution and imprint credibility. The right choice depends on what you want the book to achieve and how much of the work you're willing to fund and run.

Do I need a literary agent to get a traditional publishing deal?

For most large publishers, yes. Many don't accept unsolicited manuscripts and only consider books submitted through agents. Some smaller and independent presses do accept direct submissions, so check each publisher's guidelines. For a major deal, finding an agent who represents books like yours is usually the first real step after your manuscript or proposal is ready.

Will bookstores stock a self-published book?

Some will, but it's harder. Physical bookstores usually order through established distribution channels and prefer books with returnable stock and proven demand. Local and independent stores are often open to local authors, especially with an event attached. Most self-published authors sell mainly online and direct, through their website, talks and events, where shelf placement matters much less.

Can I self-publish first and get a traditional deal later?

Yes, and some authors do exactly that. A self-published book with strong sales, real reviews and an engaged audience gives agents and publishers evidence that readers want your work. Be aware that a publisher may want the rights to the existing book or may prefer a new title instead, so keep your options open and track your results carefully.

Meanwhile, email quietly returns more per dollar than almost any channel in existence, with Litmus putting the average at $36 back for every dollar spent, and unlike your social following, you actually own it.

Why is email the highest-ROI channel? Because it reaches people who already asked to hear from you, it costs next to nothing per send, and no platform sits between you and the inbox deciding who sees your message. Low cost, warm audience, direct access. That combination is almost impossible to beat anywhere else in marketing.

The channel you own vs. the channels you rent

Your social followers are not yours. They are the platform's, loaned to you at a reach the platform decides. Change the algorithm and your audience evaporates overnight. Your email list is different. It is an asset you control, a direct line to people who already raised their hand, with no gatekeeper between you and them.

That ownership has practical consequences. You can export your list and move software providers tomorrow. You can email everyone on it, not the fraction an algorithm chooses. You can tie every subscriber to what they bought and what they care about. Try doing any of that with your Instagram followers. The deeper case for owning your audience is laid out in why renting your audience means losing it.

Building your business on rented land feels free, right up until the landlord changes the locks.

Why email prints money

• It reaches people who already said yes, warm, not cold.
• It costs almost nothing to send, so nearly all of what it earns is profit.
• It works while you sleep through automated sequences that welcome, nurture, and sell on autopilot.
• It compounds, every subscriber you add today keeps earning for years.
• It is measurable to the cent, so you know exactly what is working.

Look at that list next to paid social. With ads, you pay every single time you want someone's attention, and you pay more when competition heats up. With email, you pay once to acquire the subscriber, then talk to them as often as they'll welcome it for the cost of your software. That's why the businesses who take email seriously see it carry more of their revenue every year. The asset gets bigger while the cost barely moves.

Most brands do email wrong

They blast the same discount to everyone, once a quarter, then wonder why nobody opens. Email that works is a relationship, not a megaphone. It welcomes new subscribers properly, segments by interest and behaviour, tells stories, delivers value, and earns the right to sell, so that when you do ask for the sale, people are ready.

The quarterly blast fails for a simple reason. It trains your list that every email is a sales pitch. So they stop opening. Then inbox providers notice nobody opens your emails and start filtering them into promotions or spam. Then the next blast reaches even fewer people. Neglect doesn't just waste the list. It actively damages it.

The five flows every business should have

Forget the newsletter for a moment. The real money in email lives in automated flows: sequences that trigger on something a subscriber does and then run on their own, forever. Build these five and you have an email system, not an email habit.

1. The welcome sequence. Three to five emails over the first week or two after someone joins. Deliver what you promised, tell your story, show proof, answer the obvious objections and make a clear first offer. New subscribers are at peak interest. This is where many businesses leave the most money on the table.
2. The nurture sequence. For leads who aren't ready to buy. Useful, specific emails that solve problems and build trust, each with a soft path to buying when they're ready.
3. The abandoned action sequence. Abandoned carts for ecommerce. Abandoned quote requests or booking forms for service businesses. A short, helpful reminder catches people who were one step from yes.
4. The post-purchase sequence. Thank them, help them get the most from what they bought, ask for a review, then introduce the next logical purchase. The easiest customer to sell to is one who just bought.
5. The win-back sequence. For customers who haven't bought or opened in a while. A genuine check-in, a reason to come back and, if they don't respond, a polite goodbye that cleans them off your list.

Once these are written, they work every day without you touching them. Your weekly or monthly broadcast emails then sit on top as the relationship layer: stories, tips, news and the occasional offer.

If you only have time for one, build the welcome sequence first. Every other flow depends on subscribers who trust you, and the welcome sequence is where that trust gets set. Write it once, in your own voice, and it greets every new subscriber you ever get. Then add the post-purchase flow, because it turns one sale into a second one with almost no extra effort.

A worked example: the florist with 2,000 forgotten emails

Say you own a florist in Adelaide. Over the years you've collected about 2,000 email addresses from online orders. You email them twice a year, before Valentine's Day and Mother's Day, with a discount code. Open rates are poor and you've half written off email as a channel that doesn't work for you.

Now look at what's actually sitting in that list. You know who bought, when, and what the occasion was. Somebody who ordered anniversary flowers on the 12th of March last year is very likely to need anniversary flowers around the 12th of March this year. Somebody who ordered sympathy flowers probably doesn't want a cheerful discount blast next week.

So you build a simple system. A welcome sequence for new buyers that tells your story and shows how you choose and arrange your stems. A reminder flow that emails people two weeks before the anniversary of a past order: "Last year you sent roses on the 12th. Want us to have something ready again?" A post-purchase email asking for a review. A monthly broadcast with seasonal flowers and care tips, and no discount.

You haven't bought a single new subscriber. You've just started using what you already owned. That's the kind of result email is quietly capable of when it's treated as a system rather than a megaphone.

How to measure email properly

Email is measurable to the cent, but only if you measure the right things. Open rates have become less reliable as a headline number because some email apps, such as Apple Mail with Mail Privacy Protection, load images automatically, which can register opens that didn't really happen. Treat opens as a rough health signal, not a scorecard.

Focus on these instead:

• Revenue per subscriber. Total email revenue divided by list size over a period. This is the single best number for knowing if your list is an asset.
• Revenue per flow. Which automated sequences actually make money. Improve the winners and fix or kill the rest.
• Click rate. Clicks are a real action and a stronger signal of interest than opens.
• Unsubscribes and spam complaints. A spike tells you an email missed the mark. Complaints especially can hurt your deliverability.
• List growth from quality sources. New subscribers from your pricing page or a useful guide are worth more than subscribers from a giveaway.

If you're reporting on vanity numbers elsewhere in your marketing, email is where you can break the habit. We explain why chasing the wrong numbers costs so much in vanity metrics are quietly bankrupting your marketing.

The mistakes that quietly kill email

• Buying lists. People who never asked to hear from you will ignore you, mark you as spam and wreck your sender reputation. Anti-spam rules in Australia and many other markets also require consent.
• Only emailing when you want something. If every message is a sale, your list learns to ignore sales.
• Sending everything to everyone. A new subscriber and a five-year customer need different messages. Even basic segmentation by buyer versus non-buyer makes a noticeable difference.
• Ignoring the technical setup. Your sending domain needs proper authentication records set up so inbox providers trust your email. It's a one-time job that many businesses never do.
• Never cleaning the list. Subscribers who haven't opened or clicked in a long time drag down your engagement. Win them back or let them go.

Email in a world of AI and algorithms

As social reach gets less predictable and more search gets answered by AI before anyone clicks, the channels you control get more valuable, not less. A subscriber list is a direct line that no algorithm update or AI summary can intercept. It's also the place where your best ideas get read in full, by people who chose to hear them.

Email also feeds everything else. Your content engine gets distributed through it, a point we make in building a content engine, not a content calendar. Your ad platforms perform better when you upload real customer lists. Your launches start with warm buyers instead of cold traffic.

Stop sitting on money

We build email systems that run quietly in the background and pay for themselves many times over, welcome flows, nurture sequences, launch campaigns, win-backs. The unglamorous machine that serious brands rely on while their competitors chase vanity. If you have a list gathering dust, you are sitting on money. Let's go collect it.

It's part of our wider digital marketing work, and it's often the first thing we look at, because it's the fastest place to find revenue a business already owns. Start with one flow this week. The welcome sequence is the best place to begin.

Frequently asked questions

Is email marketing still worth it for small businesses?

Yes. Email gives a small business direct access to people who have already shown interest, without paying for every impression the way you do with ads. It costs little to run, you own the list outright and automated flows keep selling without extra effort. For most small businesses, a well-built welcome and post-purchase sequence is one of the highest-return marketing jobs they can do.

How often should a business send marketing emails?

Often enough to stay familiar, never so often that you run out of useful things to say. For most small businesses that's somewhere between weekly and monthly broadcasts, with automated flows running in the background. Consistency matters more than frequency. A list that hears from you regularly with genuinely useful content will respond far better to offers than one you only contact when there's a sale.

What is a welcome email sequence?

A welcome sequence is a short series of automated emails that goes to every new subscriber, usually three to five messages over one to two weeks. It delivers what they signed up for, introduces your story, shows proof you can help and makes a clear first offer. New subscribers are at their most interested, so this sequence often does more selling than any broadcast email.

Why are my marketing emails going to spam?

Common causes include missing email authentication records on your sending domain, low engagement from a list that rarely opens, purchased or old addresses, and spam complaints from people who don't remember signing up. Fix the technical setup first, then clean out long-inactive subscribers, send content people actually want and make unsubscribing easy. Healthy engagement is what inbox providers reward.

So does your brand need a sound? If you run video, ads, a podcast, a phone line or an app, yes. Sonic branding is the deliberate, repeated use of your own audio signature so people recognise you with their ears before they read a single word. You already make sound in all of those places. The only question is whether it sounds like you or like everyone else.

Sound bypasses the rational brain

Sound carries emotion without asking anyone to look. A signature sound can attach feeling to your brand before the viewer consciously processes the logo. In a feed-scrolling, sound-on world, that's a superpower.

Think about how people actually meet your brand now. Half glancing at a phone while the kettle boils. Listening to a podcast on the drive. Hearing a reel play from the next room. In every one of those moments the eyes are busy somewhere else and the ears are wide open. Your logo can't do any work when nobody is looking at the screen. Your sound can.

There's a second advantage people miss. Sound is time-based. A visual can be skimmed in a fraction of a second, but a two-second audio signature has to be heard all the way through. For those two seconds you have the listener's full attention, whether they meant to give it to you or not.

You spend months perfecting how your brand looks. How much have you spent on how it sounds?

The layers of a sonic identity

A sonic identity isn't one jingle. It's a small system, the same way your visual identity is more than a logo. It has three layers.

• The audio logo: A short, ownable signature, two to three seconds that say "this is us." It closes your ads, opens your videos and plays when your app loads.
• The brand theme: A flexible musical bed for ads, content, and events. Built so it can stretch to 60 seconds for a brand film or shrink to 15 for a social cut without losing its character.
• The texture: The instruments, tempo, and tone that define your sonic personality across everything. This is the part that lets a composer, an editor or a podcast producer make new audio that still sounds like you, even when the logo isn't in it.

Get the three layers working together and something useful happens. Every new piece of content reinforces the same memory instead of starting from scratch. The audio logo is the handshake. The theme is the conversation. The texture is the accent you speak in.

Why stock music can't do this job

Here's where most brands trip. They know their videos need music, so they grab a track from a library, and the library track is fine. Upbeat. Inoffensive. Licensed.

The problem is that the same track is licensed to everyone. Your competitor can use it. A dropshipper can use it. A real estate agent three suburbs over probably already has. When your audio is shared property, it can't carry your identity, because it carries everyone's. We've unpacked this properly in why royalty-free music quietly erodes your brand, and the short version is this: rented sound builds somebody else's memory.

An original composition is different in three ways that matter commercially.

• It's exclusive. Nobody else can play it, so every play builds recognition for you alone.
• It's designed. It's written to express your brand's personality, not a generic mood like "corporate uplifting."
• It's modular. It can be delivered as stems and variations, so it fits a 6-second bumper, a podcast intro and a trade show loop without sounding stitched together.

How a sonic identity gets built, step by step

The process isn't mysterious. It's closer to a brand strategy job than a songwriting session. Here's the sequence that works.

1. Start with the brand, not the music. Pull out your positioning, your audience and three to five personality words. "Warm, precise, local" leads to very different sounds than "bold, fast, irreverent." If your positioning is fuzzy, fix that first, because music amplifies whatever it's pointed at.
2. Audit what you already sound like. Collect every place your brand makes a noise. Ads, reels, hold music, podcast intros, app sounds, event playlists. Most businesses find five different moods and zero consistency.
3. Build a sonic moodboard. Reference tracks, instruments, tempos and textures that fit the personality words. This is where you agree on direction before anyone writes a note, which saves expensive rewrites later.
4. Compose the audio logo. Short, simple, singable. The test is whether someone could hum it back after hearing it a few times. Complexity feels impressive in the studio and disappears in the real world.
5. Extend it into a theme and textures. Develop the logo's melody or motif into longer pieces, then create variations: an energetic cut, a calm cut, a stripped back version for voiceovers.
6. Test it on real devices. Phone speakers, car stereos, laptop speakers and cheap earbuds. If the signature only works on studio monitors, it doesn't work.
7. Document and deliver. Hand over files in the formats your team and partners actually need, plus simple guidelines on where each asset goes and where it doesn't.

How long does that take? For most small and mid-sized brands, weeks rather than months. The heavy lifting is the thinking in steps one to three. Once the direction is locked, composition moves fast.

A worked example: the clinic that sounded like everyone

Say you run a Brisbane physiotherapy group with four clinics. You post two reels a week, run Meta ads, record a monthly podcast for patients and have a phone line with hold music. Every one of those uses different audio. The reels use whatever's trending. The ads use a library track. The podcast has a free intro from a template. The hold music came with the phone system.

From the listener's side, there's no pattern to remember. Four channels, four sounds, zero recognition.

Now picture the fix. You settle on three personality words: calm, capable, encouraging. A composer writes a two-second audio logo built from a gentle rising motif, the musical equivalent of "you're getting better." That motif becomes the tail of every ad and reel. A 30-second theme built from the same motif opens the podcast. A soft, looping version replaces the hold music.

Six months later, a patient hears that rising phrase from a reel playing on their partner's phone and knows who it is before they see the screen. That's the whole point. Nothing about the clinic's service changed. It just stopped being acoustically anonymous.

Consistency makes it stick

A sound becomes a memory only through repetition. Use it on every video, every ad, every touchpoint. The brands with famous sounds didn't get there because the sound was clever. They got there because they refused to stop playing it.

Here's where your sonic identity should show up, at minimum:

• Paid ads: The audio logo on the end of every video ad, every time. This is your highest-frequency exposure.
• Organic video and reels: The logo or a theme variation, even when you're also using a trending sound. Layer it in at the end.
• Podcasts and webinars: Theme on the intro and outro. It's the easiest win there is, because listeners hear it every single episode. If your show still sounds rough, fix the podcast audio basics at the same time.
• Phone and hold music: An underrated one. People on hold are a captive audience who are already your customers.
• Apps and products: Notification sounds, startup sounds, confirmation chimes.
• Events and retail spaces: Theme variations as ambient loops, so the physical experience matches the online one.

The rule is simple. Every time you make a sound and it isn't your sound, you've spent attention and banked nothing.

The mistakes that waste the investment

Most failed sonic branding doesn't fail because the music was bad. It fails for boring, avoidable reasons.

• Changing it too soon. Your team hears the audio logo hundreds of times and gets sick of it long before customers have heard it enough to remember it. Boredom inside the business is a sign it's working, not a sign it needs replacing.
• Making it too long or too complex. If it takes eight seconds and a key change, it won't survive a 15-second ad.
• Copying a trend. Writing something that sounds like this year's viral track dates your brand the moment the trend moves on.
• Launching it quietly. Commissioning a sonic identity and then using it in two videos is like printing business cards and leaving them in the box.
• Ignoring the rights. Make sure you own or exclusively license what you paid for, in writing, across every channel you'll use it in.

How to tell if it's working

Sonic branding is a memory play, so measure memory, not just clicks. A few practical ways to do it without a research budget:

• Ask. Add "how did you first hear about us?" to your enquiry form or onboarding call, and listen for people mentioning your ads or content by their sound.
• Watch branded search. Rising searches for your brand name over the months you run the sound consistently are a signal that recognition is building.
• Test ad variants. Run the same video ad with and without the audio logo and compare results over a meaningful period. You're looking for a trend, not a miracle overnight.
• Try the hum test. Play the first second of your audio logo to regular customers and see who finishes it. Informal, but revealing.

For the bigger commercial case, including why familiar audio makes people more comfortable buying, read how audio branding drives sales.

Your brand already has a sound

This is the part most founders haven't considered. You're not choosing between having a sound and not having one. Every ad, reel and hold tone you've ever published has already defined how your brand sounds. If you didn't design it, the default is random, borrowed and forgettable.

We compose original, ownable audio identities, not stock loops anyone can license, so your brand has a voice the moment the speakers turn on. Through our music and sound team, that means strategy first, then composition, then a full kit of assets your editors can actually use. If your brand is working hard to look distinct and then sounding exactly like everyone else, that's the cheapest gap in your marketing to close.

Frequently asked questions

What is sonic branding?

Sonic branding is the strategic use of original sound to make a brand recognisable by ear. It usually includes a short audio logo, a longer brand theme and a set of sonic textures such as instruments and tempo. Used consistently across ads, video, podcasts and phone lines, it builds memory the same way a visual logo does, but it works even when nobody is looking at the screen.

How long should an audio logo be?

Many well known audio logos run about two to three seconds. That's long enough to carry a distinct melody or motif and short enough to sit at the end of a 6 or 15-second ad without eating the message. If it can't be hummed back after a few listens, it's probably too long or too complex to become a real memory.

Can a small business afford sonic branding?

Yes. A small business doesn't need an orchestra or a national campaign. It needs a simple, ownable audio logo, a flexible theme and clear rules for using them. The cost is usually modest next to a video shoot or a website, and because the assets get reused in every ad and video for years, the value per play keeps improving the longer you use it.

Is royalty-free music okay for brand videos?

Royalty-free music is fine for one-off content, but it can't build brand recognition because anyone can license the same track, including your competitors. For recurring touchpoints like ad endings, podcast intros and brand films, an original, exclusive sound works harder, because every play builds memory for your brand alone instead of for a library track shared by thousands of other businesses.

The platforms got smart. Their machine learning now finds your buyers better than any manual targeting ever could, but only if you stop tying its hands.

So does broad targeting really work better than detailed targeting? For most advertisers on Meta and Google in 2026, yes, as long as three things are in place: strong creative, clean conversion tracking and enough budget for the system to learn. Broad targeting hands the "who" to the algorithm so you can put all your effort into the "what."

You are handcuffing the algorithm

When you hyper-target, you are telling a system with billions of data points that you know better than it does who wants your product. You do not. You are shrinking its pool, starving it of signal, and forcing it to overpay for a tiny audience while it ignores buyers you never thought to include.

Think about what an interest stack actually is. It's your best guess, based on a handful of assumptions, about what your customers click on. "Women 30 to 45, interested in yoga and organic food." Meanwhile, the platform can see who actually stops scrolling, watches, taps and buys, across millions of behaviours you can't see and would never think to select.

Narrow audiences also get expensive fast. When lots of advertisers pile into the same obvious interest groups, they're all bidding for the same people. You pay more for each impression, you show the same ad to the same small crowd over and over, and your results fade. Broad gives the system room to find cheaper buyers in places nobody else is bidding.

Modern ad platforms do not need you to find the customer. They need you to feed them a great ad and get out of the way.

What "broad" actually means (and what it doesn't)

Broad targeting doesn't mean "show my ad to everyone on earth." It means removing the detailed interest and behaviour layers and keeping only the constraints that genuinely matter to your business.

• Keep: the locations you actually serve, the age limits that are legally or practically required, and the language your ads are written in.
• Remove: interest stacks, lookalike layering, job titles you guessed at and narrow demographic slices you've never tested.
• Let the platform decide: who inside that pool sees the ad, based on who responds and converts.

A local physio in Newcastle still targets people near Newcastle. What changes is that they stop trying to guess which of those locals will book, and let the system figure that out from real behaviour.

The new division of labour

• The algorithm's job: find the people most likely to buy. It is superhuman at this, let it.
• Your job: feed it a creative so strong it teaches the system exactly who responds.
• The signal: your ad's performance is the targeting. Great creative trains the machine faster than any interest list.

Go broad, hand the platform a scroll-stopping ad, and it will do something no spreadsheet of interests can: discover pockets of buyers you would never have guessed, at a cost per customer that keeps dropping as it learns.

Here's the part most people miss. Your creative is now doing the targeting for you. An ad that opens with "Tradies, your ute is costing you more than you think" will be watched by tradies and skipped by everyone else. The platform notices who engages and goes looking for more people like them. The message filters the audience. We unpack how Meta's recent changes pushed this even further in why Andromeda made 300 ads worth less than 15.

What broad targeting needs to work

Broad isn't magic. Flip it on with a weak setup and you'll burn money faster than before. Three things decide whether it works.

1. Clean conversion tracking. The system learns from the purchases, leads and bookings you report back. If your pixel is firing twice, missing events or sending thin data, it's learning from garbage. Server-side tracking and good customer data make a real difference, and we explain why in the boring number quietly deciding your ad results.
2. The right optimisation goal. Optimise for the action you actually want. If you optimise for clicks, you'll get clickers. If you optimise for purchases or qualified leads, the system hunts for buyers.
3. Enough budget and patience to learn. The algorithm needs a steady stream of conversions to find its feet. Starving a campaign or editing it every day resets the learning and keeps it guessing.

If you can't tick all three, fix those first. Broad targeting amplifies whatever you feed it, good or bad.

Which means creative is now everything

If targeting is handled by the machine, the entire game moves to the creative. The hook, the message, the offer, that is your lever now. The advertisers winning in 2026 are not the best at audience-building. They are the best at making ads people stop for.

That changes where your time and budget should go. Instead of spending hours inside audience settings, spend it on:

• Different angles, not just different colours. Test genuinely different ideas: a pain-led ad, a proof-led ad, a founder story, a demo, a customer testimonial. Each angle speaks to a different slice of buyers.
• Hooks that call out the right person. The first second decides who keeps watching, so make it speak directly to the buyer you want.
• A steady supply of fresh creative. Winning ads wear out. Plan a regular rhythm of new concepts so performance doesn't collapse when one ad fades. Our piece on why winning ads now die in about 14 days covers the refresh cycle.

A worked example: the skincare brand that stopped guessing

Here's a hypothetical. Say you run an Australian skincare brand spending $4,000 a month on Meta. You've got six ad sets, each targeting a different interest stack: "clean beauty," "yoga," "luxury skincare," "mums," and so on. Each ad set gets a small slice of budget. None of them exit learning. Your cost per purchase creeps up every month and you keep adding new interests hoping one will break through.

You restructure. One campaign, broad targeting across Australia, adults in the age range your product suits. You consolidate the budget so the system gets enough purchases to learn. Then you pour the time you used to spend on audiences into creative: five distinct concepts, each with a different angle. One for sensitive skin, one for busy parents, one showing real customer before-and-after results, one founder story, one ingredient explainer.

Within a few weeks, the reporting shows the platform leaning into audiences you'd never have targeted, and specific creatives pulling in specific kinds of buyers. The sensitive-skin ad finds one group. The founder story finds another. The creative did the segmenting.

These numbers are illustrative, not a promise. But the pattern is the point: fewer, bigger campaigns, broad audiences, more creative variety.

When narrow targeting still makes sense

Broad isn't a religion. There are times a tighter setup is the right call:

• Retargeting. Showing ads to people who visited your site, watched your videos or joined your list is still valuable, as long as the message fits where they are in their decision.
• Very small budgets. If you can only generate a handful of conversions a month, the system may not have enough data to learn well. Tighter, well-reasoned targeting can help at the start.
• Hard eligibility limits. If you legally can only sell to certain people or places, those constraints stay.
• Niche B2B offers. Selling to a narrow professional group may call for tighter targeting, though strong creative that calls out that group still does a lot of the work.

The rule of thumb: add restrictions only when you have a real business reason, not because it feels safer.

How to switch to broad without blowing up your account

1. Audit your tracking first. Confirm conversions are firing correctly and only once. Fix this before anything else.
2. Consolidate. Merge fragmented ad sets so budget isn't spread thin across audiences that compete with each other.
3. Run broad alongside your current best performer. Test it rather than switching everything overnight, and give it enough time and spend to judge fairly.
4. Load it with varied creative. Several genuinely different concepts, not five versions of the same ad with new headlines.
5. Judge on business results. Cost per purchase, cost per qualified lead, revenue and return on ad spend. Not reach, not clicks.
6. Stop fiddling. Resist daily edits during the learning phase. Change creative, not settings.

The mistakes that sink most switches are predictable. People go broad but keep running the same tired ad, so the system has nothing new to learn from. They optimise for cheap link clicks and then wonder why the traffic never buys. They panic after three days of choppy numbers and pull the plug before learning finishes. Or they run broad and narrow campaigns side by side that end up bidding against each other for the same people.

Once it's running, read the breakdown reports rather than the headline numbers. Look at which placements, ages and regions the platform is favouring, and which creatives are winning with which groups. That's free market research. It often shows you a customer segment you'd never have written into an interest stack, and that insight should flow straight back into your next round of creative.

Stop fighting the machine

We build broad-but-brilliant campaigns, letting the platform do what it does best while we obsess over the creative that trains it. If you are still micro-managing audiences and wondering why costs keep climbing, you are fighting the machine instead of feeding it. Let's change that. See how our advertising team structures campaigns for the way the platforms work now.

Frequently asked questions

What is broad targeting on Facebook and Instagram ads?

Broad targeting means running Meta ads without detailed interest or behaviour layers, keeping only essentials like location, age and language. The platform's machine learning then decides who sees each ad based on who engages and converts. It works best with strong creative, accurate conversion tracking and enough budget for the system to learn.

Is broad targeting good for small budgets?

It can be, but it's riskier. The algorithm learns from conversions, so a very small budget that produces only a few results a month gives it little to work with. Consolidate spend into one campaign, optimise for a conversion that happens often enough, and give it time. If results stay weak, reasonable targeting constraints can help early on.

How long should I test broad targeting before judging it?

Long enough for the campaign to finish its learning phase and gather a meaningful number of conversions. For most accounts that's at least a couple of weeks without major edits. Compare it against your existing best campaign on cost per purchase or qualified lead, not on early click or reach numbers, which can mislead.

If targeting is automated, what should advertisers focus on?

Creative, tracking and offer. Build a steady pipeline of genuinely different ad concepts with hooks that call out your ideal buyer, make sure conversion data flowing back to the platform is clean and complete, and make the offer compelling. Those are now the main levers a human controls, and they decide most of your results.

So what are the speed rules that actually move rankings in 2026? Three numbers, called Core Web Vitals: how fast your main content appears (LCP), how fast the page reacts when someone taps (INP), and how much the layout jumps around while it loads (CLS). Pass all three for real visitors and you have removed a handbrake on your rankings and your revenue. Fail them and every dollar you spend on content and ads is working uphill.

Here's what each number means, how Google actually measures it, and the short list of fixes that do most of the work.

The three numbers to obsess over

• LCP (Largest Contentful Paint), your main content should render in under 2.5 seconds. This is the "is anything happening?" metric.
• INP (Interaction to Next Paint), replaced FID in March 2024 and measures real responsiveness. Aim for under 200ms on every tap and click.
• CLS (Cumulative Layout Shift), keep it below 0.1 so nothing jumps under the user's finger.

Each one maps to a feeling your visitor has, whether they could name it or not. LCP is patience: "is this thing broken?" INP is control: "did my tap register?" CLS is trust: "why did I just hit the wrong button?" A site that fails any one of them feels cheap, even when the design is gorgeous.

INP is the one catching businesses out right now. The old metric it replaced only measured the delay on the very first interaction. INP watches every tap, click and keypress across the whole visit and reports one of the slowest. That means the chunky mega menu, the filter panel on your shop page and the booking widget three screens down all count now. Plenty of sites that looked "fast" under the old rules are quietly failing the new one.

How Google actually scores you

This is where most people get confused, so let's clear it up.

There are two kinds of speed data. Lab data is what you get when you run a test tool like Lighthouse: one simulated visit, on one simulated device, at one moment. Field data is what real people experience on real phones over real connections, collected from Chrome users who share usage statistics.

Google ranks on the field data. Your Lighthouse score is a diagnostic, not a grade. You can hit 100 in the lab on your office wifi and still fail in the field because half your customers are on a mid-range Android in a car park with one bar of signal.

Google also judges each metric at the 75th percentile of visits, not the average. In plain English, three out of four visits need to hit the target. Your fastest visitors don't rescue you. Your slowest quarter decides it. That's why testing on your own new laptop tells you almost nothing about how the site feels to the people paying you.

Where to look: Google Search Console has a Core Web Vitals report that groups your URLs into good, needs improvement and poor, using field data. PageSpeed Insights shows both field and lab results for a single URL. Start with Search Console, because it tells you which templates are failing across the whole site instead of one page at a time.

What actually moves these numbers

Most "optimisation" advice is noise. The high-leverage moves are boring and devastatingly effective:

Ship less JavaScript. Serve images in next-gen formats. Reserve space for everything before it loads.

Edge caching, modern image formats like WebP and AVIF, font preloading, and ruthless code-splitting do more for your rankings than any keyword trick. Speed compounds. Google says it can crawl more when a site responds quickly, and faster pages convert better, which feeds the loop again.

Here's how those moves line up against each metric, so you know which lever to pull for which problem.

Fixing a slow LCP

Your largest element is usually the hero image, a hero video poster or a big headline. Find out which one it is first, because the fix depends on it.

• Serve the hero image as WebP or AVIF, sized for the screen it's shown on. A 4000 pixel photo squeezed into a phone screen is one of the most common LCP killers on the web.
• Never lazy-load the hero. Lazy loading is for images below the fold. Tell the browser to fetch the hero early, with high priority.
• Cut the server response time. Cache pages at the edge so the HTML arrives from a data centre near your visitor, not from one server on the other side of the world.
• Stop render-blocking. Every stylesheet and script in the head that must load before the page paints is time added to LCP.

Fixing a slow INP

INP is almost always a JavaScript problem. When someone taps, the browser can't respond while it's busy running other code.

• Audit your third-party scripts. Chat widgets, heatmaps, review badges and old tracking pixels pile up for years. Each one competes for the main thread.
• Break up long tasks. Code that runs in one big chunk blocks every tap until it's done. Split it so the browser can breathe between pieces.
• Give instant feedback. Show the button pressed or the menu opening straight away, then do the heavy work after the paint.

Fixing a bad CLS

• Give every image and video explicit width and height so the browser reserves the space before it arrives.
• Reserve room for ads, embeds and cookie banners instead of letting them shove content down.
• Preload your fonts and use a fallback with similar proportions, so text doesn't reflow when the web font lands.

Your one-week Core Web Vitals audit

You don't need a developer to find out where you stand. You need an afternoon and some honesty. Here's the order to do it in.

1. Open Search Console's Core Web Vitals report. Note which URL groups are poor on mobile. Mobile first, always, because that's where most of your visitors are and where the numbers are worst.
2. Pick your three money pages. Usually the homepage, your top service or product page, and the landing page your ads point at. Run each through PageSpeed Insights.
3. Identify the failing metric per page. Don't try to fix everything. Find the one number in the red on each page.
4. Find the culprit element. PageSpeed Insights tells you which element is the LCP, which scripts are hogging the main thread and which elements shifted.
5. List every third-party script. Ask of each one: who uses this data, and when did they last look at it? Delete what nobody can defend.
6. Fix, then wait. Field data rolls over a 28-day window, so improvements show up gradually. Don't panic when nothing changes on day two.

The business case

Even tiny delays cost money. Deloitte's Milliseconds Make Millions study found a 0.1 second improvement in mobile site speed lifted retail conversions by 8.4%. When we rebuild a site for performance, we aren't chasing a green Lighthouse score for bragging rights, we are recovering revenue that was leaking out the back door the entire time.

Let's make that concrete with a hypothetical. Say you run a Brisbane physio clinic spending $4,000 a month on Google Ads, all pointed at a booking page that takes five seconds to show anything useful on a phone. Some of the people you paid to click simply leave before the page appears. You paid for the click, Google counted it, and the patient went to the clinic down the road whose page loaded first.

Now fix the page so it loads in under two seconds. You haven't changed the ads, the offer or the budget. You've just stopped losing people in the gap between the click and the content. Every visitor you keep is a customer you already paid for, finally getting the chance to buy. That's why speed work often pays for itself faster than any new campaign. It makes the money you're already spending work harder.

It also lowers what you pay for traffic in the first place. Landing page experience feeds into how Google Ads judges your ads, and a slow, jumpy page is a poor experience by any definition. We break down the full cost of a sluggish site in why a slow website is a silent business killer.

Common mistakes that waste the budget

Chasing the lab score. Teams burn weeks turning a Lighthouse 88 into a 97 while the field data, the thing Google actually uses, doesn't move. Fix what real users experience.

Adding a speed plugin on top of a bloated build. A caching plugin on a page builder theme loaded with sliders, animation libraries and six font families is a bandage on a broken leg. Sometimes the honest answer is a rebuild, which is exactly the trap we unpack in the real cost of a cheap website.

Testing on desktop. Your desktop score is almost always better. Your customers are almost always on mobile. Test where they are.

Treating it as a one-off. Speed decays. Every new tracking pixel, every marketing embed, every "quick" homepage carousel adds weight. Without someone guarding it, a fast site gets slow again within a year.

Autoplaying heavy video in the hero. Video sells, but a full-weight file loading before anything else is a brutal LCP hit. Use a light poster image, compress hard, and host video on infrastructure built for it.

Speed is a habit, not a project

Core Web Vitals won't win you rankings on their own. Great content, clear answers and real authority still do the heavy lifting. But passing them removes a tax you're currently paying on every page, every click and every campaign.

The brands that win treat speed like hygiene. Every new feature gets asked the same question before it ships: what does this cost us in load time, and is it worth it? That single habit keeps a site fast for years.

And speed matters most on the pages where money changes hands. If your ads land on a slow page, fix that page first. We cover how to build those pages properly in the anatomy of a landing page that prints money.

If you'd rather hand the whole thing to people who build for speed from the first line of code, that's what our web development team does. We build lean, we test on real phones, and we measure what your customers actually feel, not what a lab simulator guesses.

Frequently asked questions

Do Core Web Vitals still affect Google rankings in 2026?

Yes. They're part of how Google assesses page experience, using field data from real Chrome users. They won't outrank genuinely better content on their own, but when two pages are close on relevance, the faster, more stable page has the edge. They also affect conversions directly, which matters even more than rankings.

What is a good INP score?

Aim for under 200 milliseconds, measured on real visits. INP tracks the delay between a tap, click or keypress and the next visual update, across the whole visit rather than just the first interaction. Heavy JavaScript and piles of third-party scripts are the usual cause of a poor score, so start your audit there.

Why is my PageSpeed score good but Search Console says my pages are poor?

The PageSpeed performance score is lab data from one simulated visit. Search Console uses field data from real visitors over 28 days, judged at the 75th percentile. If your slowest quarter of visitors are on older phones or weak connections, you can look fast in the lab and still fail in the field. Google ranks on the field.

How long does it take to see Core Web Vitals improvements?

Field data is collected over a rolling 28-day window, so after you ship a fix the reported numbers improve gradually over about a month. Your conversion rate can respond much faster, because real visitors feel the faster page from the first day. Measure both, and give each change time before judging it.

Change your traffic and you get a little more of the same. Change your offer and you can change everything. A great product with a weak offer struggles. An ordinary product with an irresistible offer can dominate a market.

So how do you make buying from you a no-brainer? You stack the value until the price looks small, remove the risk until saying yes feels safe, make the outcome specific, and give a genuine reason to act now. Do all four and your prospect stops comparing you with competitors, because nobody else is offering anything close.

What makes an offer irresistible

An offer is not a price tag. It is the entire equation running in your prospect's head: what do I get, what does it cost me, what could go wrong, and is the reward worth the risk? Your job is to tilt that equation so far in their favour that saying no feels like the stupid move. That is a Godfather Offer, so good they cannot refuse it.

People do not buy products. They buy the belief that the reward is worth more than the risk. Stack the reward, kill the risk.

Notice that "cost" in that equation is bigger than money. Your prospect is also weighing time, effort, embarrassment if it goes wrong, the hassle of switching from what they use now, and the fear of making a decision their boss or partner will question. Most businesses only compete on the money part. The ones who win reduce every other cost too.

Why your current offer is probably weak

Here's a quick way to tell. Read your offer out loud as a prospect would hear it. If it sounds like "we do X for a fee", you don't have an offer. You have a service description with a price attached.

Weak offers share the same symptoms:

• Prospects ask "how much?" first. When the value isn't obvious, price is the only thing left to compare.
• You keep hearing "let me think about it". That's code for "the risk still feels bigger than the reward".
• Discounting is your main closing tool. If the only way to win the sale is cutting your margin, the offer isn't carrying its weight.
• You sound like every competitor. Same deliverables, same process, same "quality service". The prospect can't see a difference, so they pick the cheapest.

If that last point stings, it's worth reading why "being the best" is quietly killing your business. Claiming to be better is weak. Being clearly different is strong.

The levers that build one

• Stack the value until the price looks small next to what they receive.
• Remove the risk with a guarantee bold enough to make you slightly nervous.
• Add a reason to act now, real scarcity or urgency, not fake countdowns.
• Make it specific. A concrete outcome beats a vague promise every time.
• Solve the whole problem, not just a slice, so there is nothing left to hesitate over.

Let's go deeper on each, because the details decide whether an offer is irresistible or merely decent.

Stacking value the right way

Value stacking isn't piling on random bonuses. Every item in the stack should remove an obstacle between your customer and the result they want. Ask: what will they struggle with after buying? What will they need next? What slows them down? Then build a piece of the offer that solves each one. A template, a checklist, a setup session, priority support, a done-for-you step they'd otherwise do badly themselves. Each should feel like a relief, not filler.

Guarantees that make you slightly nervous

A guarantee shifts risk from the buyer to you. That's the point. "Satisfaction guaranteed" means nothing because everyone says it. A strong guarantee is specific: what result, by when, and what you'll do if it doesn't happen. Refund, redo the work free, keep working until it's delivered. Pick the one you can honour and your competitors won't match. If you're genuinely good at what you do, the few people who claim it cost far less than the extra sales it wins.

Urgency that's actually real

Fake countdown timers that reset on refresh destroy trust. Real reasons to act now are everywhere if you look: limited capacity each month, a price rise on a set date, a seasonal deadline, a bonus tied to a launch window. Tell the truth about why the offer changes, and hold to it.

Specific beats vague, every time

"Grow your business" is a wish. "A new website live in 21 days, built to turn your ad clicks into booked calls" is a promise someone can picture. Specific outcomes, timeframes and deliverables make the reward feel real, which makes it easier to believe, which makes it easier to buy.

Solving the whole problem

Customers don't want a slice of a solution. They want the problem gone. If you build the website but leave them to write the copy, source the photos and set up the tracking, you've handed them three new problems and called it a service. Every gap you leave is a reason to hesitate, and a reason to shop around for someone who covers it. Map the full journey from "I have this problem" to "this problem is solved", then own as much of it as you can deliver well. The more complete the solution, the less there is to compare.

How to build your Godfather Offer this week

Here's a step-by-step process you can work through with a notepad and a couple of hours.

1. Define the dream outcome. Not what you do, what the customer gets. Write it in their words, as specifically as you can.
2. List every obstacle. Every reason a customer might fail to reach that outcome, or might hesitate to buy. Aim for twenty or more.
3. Turn each obstacle into a solution. For every item on the list, write one thing you could include, do or promise to remove it.
4. Trim to the high-value, low-cost pieces. Keep the solutions customers would value highly that cost you relatively little to deliver. Cut the rest.
5. Add the guarantee. Choose the boldest one you can genuinely honour.
6. Add the real reason to act now. Capacity, deadline, or price change. It has to be true.
7. Name it. Give the package a name that describes the outcome, so it stops sounding like a generic service.
8. Test it on real prospects. Watch their faces and listen to their questions. Every objection that still comes up tells you what to fix next.

A worked example: the accountant who stopped selling tax returns

Here's a hypothetical. Say you run a small accounting practice in Adelaide. Your current offer is "business tax returns from $1,500". Every prospect compares you on price with three other firms and the cheapest one usually wins.

Now rebuild it with the levers. The dream outcome for a busy tradie owner isn't a tax return. It's never thinking about tax, never getting a nasty surprise bill, and knowing every month whether the business is actually making money.

So the new offer becomes a fixed monthly package: quarterly tax planning so there are no surprises, a one-page profit report each month in plain English, a direct line to one named accountant, and the annual return included. The guarantee: if the tax planning is wrong and you get an unexpected bill, the next quarter is on the house. The urgency: you only take on a limited number of new clients each quarter, because that's genuinely how many one accountant can serve well.

Same accountant. Same skills. But the prospect is no longer comparing tax return prices. They're comparing "stress-free finances with a guarantee" against "a tax return". That's not a close call, and the price conversation changes completely. For more on why that works, see the psychology of premium pricing.

Common mistakes when building an offer

Adding bonuses nobody wants. A stack of irrelevant extras makes the offer look desperate. Every item must solve a real obstacle.

Promising what you can't deliver. A guarantee you can't honour is a liability, not an offer. Be bold, but be able to keep your word. It's also worth checking your guarantee against Australian Consumer Law, which sits alongside anything you promise, not instead of it.

Burying the offer. A brilliant offer hidden on page three of your website does nothing. Lead with it everywhere: ads, homepage, landing pages and sales calls. A dedicated page built around one offer is where this really pays, which is why landing pages that print money start with the offer, not the design.

Making it complicated. If a prospect can't repeat your offer to their partner in one breath, simplify it.

Why this changes the whole business

When your offer is irresistible, everything downstream gets easier. Your ads convert better because the offer does the heavy lifting. Your sales calls shorten because the value is obvious. Your referrals rise because customers feel like they got away with something. One great offer lifts every metric you have.

How do you know it's working? Watch the leading indicators. Conversion rate from enquiry to sale. How often prospects ask about price before anything else. How many "let me think about it" replies you get. How often you're forced to discount. A stronger offer moves all of them in the right direction, usually faster than any change to traffic or ad spend.

Fix the lever that matters most

We engineer offers that make buying feel like the obvious, easy, low-risk choice, the kind of offer that quietly doubles a business without touching the product. If you are working on your traffic before you have fixed your offer, you are pushing on the wrong lever. Let's fix the one that matters most.

It's a core part of our consulting work at LIQID NOISE, because every other piece of marketing performs better once the offer is right. Fix the offer first, then spend on getting it seen.

Frequently asked questions

What is a Godfather Offer?

A Godfather Offer is an offer so good the customer feels foolish saying no. It stacks enough value that the price looks small, removes the risk with a strong guarantee, promises a specific outcome, and gives a genuine reason to act now. The name comes from the film line about making someone an offer they can't refuse.

How do I make my offer more compelling without lowering my price?

Increase the value and reduce the risk instead of cutting the price. Add pieces that remove obstacles to the customer's result, make the outcome more specific, add a guarantee you can honour, and shorten the time to results. Customers pay more willingly when the reward clearly outweighs what they're giving up.

What kind of guarantee should a service business offer?

Pick a specific guarantee tied to an outcome you control. Examples include redoing the work free until it meets an agreed standard, refunding if a deadline is missed, or continuing to work at no charge until a defined result is delivered. It should feel slightly uncomfortable for you and must be something you can always honour.

Is fake urgency bad for sales?

Yes. Countdown timers that reset and "last chance" offers that never end teach customers not to trust you, and misleading claims can breach consumer law. Emma Sleep was ordered to pay $15 million in penalties over misleading sale claims, which included a countdown timer that reset during a sale. Use real urgency instead: genuine capacity limits, a price increase on a stated date, or a seasonal deadline. Honest urgency still works, and it protects your reputation.

So how does audio branding drive sales? It builds recognition and emotion in a way visuals alone often can't. A consistent sound tells people it is you before they read a word, attaches a feeling to your brand, and makes every ad, video and touchpoint reinforce the same memory. Familiar brands get chosen more easily, and sound is one of the fastest routes to familiar.

You do not need a global budget to do this. You need a clear idea of how you want to be felt, and the discipline to sound that way every single time.

Sound skips the bouncer

Your rational mind is a bouncer, it scrutinises claims, resists sales pitches, second-guesses logos. Sound walks right past it. Music tends to hit feeling first and reasoning second, and a familiar tune can stay with people for years. That is why a three-second sound can trigger instant recognition long after you last heard it.

Think about how you actually consume content. You look away from the screen. You scroll with your thumb while half listening. You have a video playing in another tab. Your eyes are busy, divided and easily distracted. Your ears are always on. A sound can reach someone who is not looking at you at all, and that is a reach no logo can match.

A logo asks to be looked at. A sound gets remembered whether you wanted it to or not.

What audio branding actually includes

Audio branding is bigger than a jingle. Think of it as a sonic identity system, the audio equivalent of your visual brand guidelines. A complete one usually covers:

• The sonic logo. A short signature, often just a few notes or a distinctive sound, that closes ads, opens videos and marks your content.
• The brand theme. A longer piece of music built from the same melodic or tonal DNA, which can be rearranged for different moods and lengths.
• Sound palette. The instruments, textures and production style that belong to you. Warm acoustic guitar and soft piano say something very different from punchy synths and heavy drums.
• Voice guidelines. The tone, pace and character of any voice speaking for your brand in ads, videos or on hold.
• Functional sounds. Notification chimes, app interactions, confirmation tones. Small moments that get heard hundreds of times.

For a deeper look at why a brand needs this system in the first place, read our piece on sonic branding and why your brand needs a sound.

Where audio branding earns its keep

• A sonic signature, a short, ownable sound that marks every video, ad, and touchpoint as unmistakably you.
• Consistent scoring across your content so your brand feels the same everywhere.
• Emotional targeting, music chosen to make people feel exactly what you want them to feel before you say a word.
• Recognition at speed, so people know it is you in the first second of a scroll.

Each of these pays off in a specific place. Your paid social ads, where the first second decides whether someone keeps watching. Your YouTube pre-roll and podcast sponsorships, where your audience is listening more than looking. Your reels and short-form content, where a recurring sound turns a pile of separate posts into a recognisable series. Your showroom, events and phone line, where sound shapes how a physical or live experience feels. Everywhere your brand makes a noise, it is either building memory or wasting the chance.

Why generic music quietly costs you

Most small brands grab a track from a stock library, drop it under a video and move on. It feels harmless. It is not. That same track is under thousands of other videos, often from brands in your own category. Every time your audience hears it, they are reminded of nobody in particular. You are paying, in attention, to build someone else's blur.

We argue the full case in royalty-free music is quietly killing your brand. The short version: borrowed sound cannot become your sound. Only original, consistent audio can do that job.

The unfair advantage hiding in plain sight

Here is why this is such an opportunity: your competitors are obsessing over how they look and completely ignoring how they sound. The entire audio dimension of branding is sitting there, unclaimed, in most markets. The brand that owns a sound owns a shortcut to memory that no visual redesign can match.

There is a second advantage. Visual trends date fast. Colours, fonts and layouts that looked fresh three years ago now look tired, and brands redesign to keep up. A well-built sonic identity ages far more slowly, especially when it is rooted in a simple, strong melodic idea. Build it once, properly, and it can carry your brand for years while it keeps compounding recognition.

How to build a sonic identity, step by step

1. Define the feeling. Pick three to five words for how your brand should feel. Confident, warm, playful, precise, rebellious. These become the creative brief for every sound decision.
2. Audit what you already sound like. Gather every ad, video, podcast intro and on-hold message. Listen back to back. Most brands discover they sound like five different companies.
3. Listen to your category. Note what competitors use. If everyone in your space uses the same uplifting acoustic stock track, that is your cue to sound different.
4. Compose the core idea. A short melodic or tonal motif that is simple enough to recognise in a second and flexible enough to live in many arrangements.
5. Build the system around it. A sonic logo, a full brand theme, shorter cutdowns, and alternative moods such as upbeat, reflective and urgent, all built from the same DNA.
6. Write the guidelines. When to use which version, which instruments belong to you, what voice style to cast, what never to do.
7. Roll it out everywhere, then repeat relentlessly. The power comes from repetition. The moment you swap it out because you are bored of it is usually right before your audience starts recognising it.

What does it take? Less than most founders expect, and more than a stock library subscription. The real investment is in thinking before composing: getting the feeling right, hearing the category, and agreeing internally on what the brand should never sound like. The composition and production then follow a brief instead of a guess. Scope depends on how many versions you need. A sonic logo and one theme is a sensible starting point for a small brand. A business with an app, a podcast, a phone line and physical venues needs the fuller system. Either way, the cost is paid once and the recognition keeps building every time the sound plays. Compare that with ad spend, which stops working the moment you stop paying, and the maths of owning a sound starts to look very different.

A worked example: the regional gym chain

Say you run a hypothetical chain of four gyms across regional Victoria. Your ads, reels and class videos each use whatever trending track the person editing them liked that week. Nothing sounds like anything.

You commission a sonic identity. The brief: energetic, welcoming, local, never intimidating. The result is a four-note motif built on punchy but friendly drums and a bright synth line. It becomes a two-second sonic logo at the end of every ad, a thirty-second theme for class promo videos, a stripped-back version for member testimonials, and a gentle chime version for your booking app.

Six months in, members start humming it in the change rooms. People in town recognise your ads before your logo appears. Your testimonial videos feel like part of one story instead of random posts. None of that required a bigger ad budget. It required choosing a sound and sticking with it.

That outcome is illustrative, not a promise. But it shows the mechanism: consistency turns small, repeated moments into a memory that belongs to you.

Common audio branding mistakes

• Too complicated. If it cannot be hummed or recognised in a moment, it will not stick.
• Chasing trends. A sound built on this year's production fad dates as fast as the fad.
• Inconsistent use. A sonic logo used on one ad in five builds almost nothing. Repetition is the whole game.
• Ignoring audio quality. A great sound played through thin, badly mixed audio still reads as cheap. The same problem sinks podcasts, as we cover in why your podcast sounds amateur.
• Mismatched feeling. A playful jingle on a funeral service or a brooding score on a kids' brand creates confusion, not memory.

How to tell if it is working

Audio branding is a long game, but it is not unmeasurable. Watch whether your video hold rates improve once the sonic logo is consistent. Ask new customers in onboarding whether they remember your ads and what they remember about them. Run simple recall tests with your audience: play the sound without the visuals and see who names you. Track branded search and direct traffic over time. None of these alone proves the case, but together they show whether your sound is starting to belong to you.

Sound as good as you look

We craft sonic identities and original scoring that make brands instantly recognisable and impossible to forget, the same weapon the giants use, built for ambitious brands ready to sound as good as they look. Our music and sound team handles the whole system, from the first motif to the final guidelines. If your brand has no sound, you are leaving a sense, and a memory, on the table. Let's claim it.

Frequently asked questions

What is audio branding?

Audio branding is the strategic use of sound to identify and express a brand. It includes a sonic logo, a brand theme, a consistent sound palette, voice guidelines and functional sounds like notifications. Used consistently across ads, videos, apps and physical spaces, it helps people recognise your brand faster and attach a specific feeling to it.

Is a sonic logo worth it for a small business?

Often, yes, especially if you publish a lot of video, run paid social or advertise on audio platforms. A small business has fewer chances to get noticed, so making every exposure recognisable matters more, not less. The key is committing to it. A sonic logo only pays off when it is used consistently across everything you publish over a long period.

How long should a sonic logo be?

Short enough to recognise almost instantly and fit at the end of any ad without slowing it down. Most work best as a brief motif of a few notes or a single distinctive sound. The simpler it is, the easier it is to remember, hum and rearrange for different moods. Length matters less than clarity and consistent use.

Can I use royalty-free music for my brand instead?

You can use it for one-off content, but it cannot become your brand's sound because anyone can license the same track. Your audience may hear it in a competitor's ad the same day. If you want sound to build recognition, it needs to be original, owned by you and used consistently. Stock music fills silence. It does not build memory.

Here's the short version. A content calendar asks "what do we post on Tuesday?" A content engine asks "what one big idea do we capture this month, and how many useful pieces can we cut from it?" The calendar burns you out chasing new ideas every day. The engine takes one substantial idea, turns it into twenty channel-native assets, and measures which ideas actually bring in leads so the next month is smarter than the last.

If you have ever stared at an empty content calendar at 4pm on a Thursday wondering what to post tomorrow, this is for you. Let's build the thing that ends that feeling for good.

Why the content calendar keeps failing you

The calendar isn't evil. It's just the wrong starting point. It treats every post as a separate job, which means every post needs its own idea, its own script, its own design and its own approval. Multiply that by five platforms and three posts a week and you have built yourself an unpaid second job.

And here's what the calendar never tells you: whether any of it worked. A filled calendar feels like progress. Green ticks, scheduled posts, a tidy grid. But a full calendar is a measure of effort, not results. Plenty of brands post every single day for a year and have nothing to show for it except a tired marketing manager.

The three symptoms of a calendar-first operation:

• Idea famine. Every week starts from zero. Nobody knows what to say, so you fall back on "happy Friday" posts and industry news reposts that build nothing.
• Shallow content. When you need a new idea every day, no single idea gets the depth it deserves. Everything is a thin slice.
• No memory. Posts go out, get whatever reach they get, and vanish. Nothing is captured, nothing is reused, nothing compounds.

One pillar, twenty assets

Start with a single substantial idea, a long-form piece, a podcast, a recorded strategy session. That becomes your pillar. From it you cut:

• Short vertical videos for Reels, Shorts, and TikTok
• Carousel breakdowns for Instagram and LinkedIn
• A quote-graphic series for the key one-liners
• An email that frames the takeaway for your list
• A blog post optimised for search

The pillar is where the thinking happens. Everything else is packaging. That flips the economics of content completely. Instead of paying the "blank page tax" twenty times a month, you pay it once, properly, and spend the rest of your time cutting, shaping and distributing.

You don't need more ideas. You need to extract more value from the ideas you already have.

What makes a good pillar (and what doesn't)

Not every idea deserves to be a pillar. A weak pillar gives you twenty weak assets, which is worse than one weak post. A strong pillar has three things going for it.

It answers a real question your buyers ask. Pull these straight from sales calls, enquiry emails and the questions your team answers on repeat. "How much does this cost?" "How long does it take?" "Why not just do it ourselves?" Those questions are gold because they sit right next to a buying decision.

It has a point of view. "Five tips for better marketing" is not a pillar. "Why your content calendar is the reason your content doesn't work" is. An opinion gives every cut-down piece a hook, because every piece is arguing something.

It is long enough to cut. A 20-minute recorded conversation or a 2,000-word article gives you material. A 90-second talking head gives you one clip. Aim for depth at the source so you have room to slice.

The easiest pillar format for most founders is a recorded conversation. Sit down with someone who asks good questions, talk for 30 to 45 minutes about one topic, and you've captured your thinking in your own voice. We cover how to squeeze a full quarter out of a single session in one video shoot, 90 days of content.

Channel-native, not copy-paste

The mistake is reposting the same crop everywhere. Each platform has its own grammar. We adapt the message to the medium, punchy and fast for TikTok, value-dense for LinkedIn, visually arresting for Instagram, so the work feels native instead of recycled.

In practice, "native" means changing more than the aspect ratio. It means:

• A new hook for every platform. The opening line that stops someone on TikTok is rarely the one that earns a click on LinkedIn. Rewrite the first line every time.
• Different lengths. A short video might carry one idea. A carousel might carry five steps. An email might carry the whole argument plus a story.
• Captions written for the scroll. Plenty of short video is watched on mute. If the point only lands with sound on, those viewers never get it.
• A clear next step that suits the channel. Follow on social, reply on email, read more on the blog, book a call on the site.

This is how small teams show up in five places at once without five times the work. If that's the goal, read how to build an omnipresent brand on a startup budget.

How to build your first engine, step by step

You don't need a studio, a team of ten or a new tool stack. You need a repeatable monthly rhythm. Here's the one we recommend starting with.

1. List your buyer questions. Spend 30 minutes writing down every question prospects ask before they buy. Rank them by how often they come up and how close they sit to a sale.
2. Pick one pillar topic for the month. Take the top question. That's your pillar. One topic, not four.
3. Capture it properly. Record a long conversation, write a deep article, or both. Get the thinking down in full, with examples and opinions.
4. Mine it for moments. Go through the transcript and highlight every strong line, story, stat you already own, and step-by-step explanation. Those highlights are your raw material.
5. Cut the assets. Turn the highlights into short videos, carousels, quote graphics, an email and a search-friendly blog post. Aim for fifteen to twenty pieces.
6. Schedule it across the month. Now the calendar has a job. It's a distribution plan for one idea, not a daily scramble for new ones.
7. Review at month end. Look at what drove enquiries, replies and sales conversations. Feed that into next month's pillar choice.

That last step is the one that turns a system into an engine. Without it, you just have a more efficient treadmill.

A worked example: one month, one pillar

Say you run a small Brisbane accounting practice. Your team gets asked the same question every week: "Should I set up a company or stay a sole trader?" That's your pillar.

In week one, you record a 40-minute conversation where a partner walks through the decision, the common mistakes and three example scenarios (all hypothetical, no client details). That single recording gives you:

• Eight short vertical clips, each answering one sub-question, like "When does a company structure start to make sense?"
• Two LinkedIn carousels: "5 signs you've outgrown sole trader" and "The mistakes people make when they switch."
• A quote-graphic series of the partner's sharpest one-liners.
• One email to your list that tells the story and invites replies.
• A long blog post that answers the question in full, built to rank and to be quoted by AI search tools.

That's roughly twenty assets from one afternoon of recording. At the end of the month, you check which clips drove profile visits, which email got replies and whether the blog post brought in any enquiries. Maybe the "mistakes" carousel pulled far more saves and comments than anything else. Great. Next month's pillar leans into mistakes.

Notice what didn't happen: nobody sat in a meeting asking "what should we post this week?"

Measure the engine, not the post

A single viral post is luck. A system that reliably produces reach, leads, and authority is a business asset. We track which pillars drive pipeline, double down on what works, and quietly retire what doesn't.

That means measuring at the pillar level, not the post level. For each pillar, ask:

• Did it bring in enquiries? Track form fills, DMs and calls that mention the topic.
• Did it grow an owned audience? Email sign-ups matter more than follows, because you own the list.
• Did it earn search visibility? Check whether the blog post starts ranking or appearing in AI answers over the following months.
• Did sales use it? The best content gets sent to prospects by your own team. If they're forwarding it, it's working.

Likes and views are useful signals for which hooks work. They are not the scoreboard. If you've been reporting on reach while revenue stays flat, read why vanity metrics are quietly bankrupting your marketing.

The mistakes that stall an engine

• Choosing pillars you find interesting instead of ones buyers ask about. Your content should answer the market's questions, not your team's hobbies.
• Cutting too thin. Twenty clips that all say the same thing is not twenty assets. Each piece should carry one distinct idea.
• Skipping the long-form written version. Video reaches people today. A deep written article keeps getting found in search and AI answers for years.
• Never reusing winners. Your best pillar from six months ago can be refreshed, re-recorded or re-cut. Most of your audience never saw it the first time.
• Letting the review slide. Without the monthly look back, you are guessing again.

Stop feeding the treadmill

The calendar asks you to be endlessly creative on demand. The engine asks you to think deeply once a month and be smart about distribution. One burns people out. The other gets better every cycle, because every month you know a little more about what your buyers actually respond to.

At LIQID NOISE, this is how we run content for brands through our digital marketing work: pillar strategy, capture, channel-native editing and pipeline tracking in one system. If your team is exhausted and your content isn't moving the needle, the answer isn't more posts. It's a better machine.

Frequently asked questions

What is a content engine in marketing?

A content engine is a repeatable system where one substantial piece of content, like a podcast episode, recorded conversation or long article, is broken down into many smaller assets for different platforms. Performance is measured at the idea level, so each month's topic choice is informed by what actually drove leads and sales. It replaces the daily scramble for new ideas with a monthly rhythm that compounds.

How many posts can you get from one piece of long-form content?

A strong 30 to 45 minute recording or a deep 2,000-word article can usually produce fifteen to twenty assets: short vertical clips, carousels, quote graphics, an email and a search-optimised blog post. The limit is how many distinct ideas the source contains. If you find yourself making clips that repeat the same point, the pillar needed more depth, not more cuts.

Is repurposing content bad for engagement?

Repurposing hurts when you paste the same crop and caption everywhere, because it feels recycled and ignores how each platform works. Done properly, it means writing a new hook, adjusting the length and format, and matching the tone of each channel. Much of your audience likely only sees you on one or two platforms anyway, so a well-adapted piece feels new to them.

How do you measure whether content marketing is working?

Measure at the topic level, not the post level. For each pillar, track enquiries that mention it, email sign-ups, search and AI visibility over the following months, and whether your sales team uses the content with prospects. Views and likes tell you which hooks work, but pipeline and revenue tell you whether the engine is paying for itself.

Why do phone videos so often outsell polished ads? Because in a social feed, polish is the fastest signal that something is an ad, and people have learned to skip ads on sight. A raw, native-looking video looks like the content people came for, so it gets watched long enough to persuade. But the winners are not careless. They are carefully built to look effortless.

The "skip" reflex

People have trained themselves to skip ads with ruthless efficiency. And the fastest way their brain identifies an ad is polish. The moment something looks like a commercial, up goes the guard and away goes the thumb. Raw, native-looking video slips under that radar because it looks like the content people actually came to watch, a real person, talking to them, not a brand performing at them.

Think about what the feed is actually full of. Friends, creators, people filming on their phones, talking straight into the lens in their car or kitchen. That is the visual language of the platform. A slow-motion drone shot with a swelling music bed and a voiceover in a deep announcer voice is speaking a different language entirely. It stands out, but for the wrong reason. It stands out as an interruption.

Once the viewer has tagged something as "ad", your hook has to fight their defences. Once they have tagged it as "content", your hook only has to be interesting. That is a massive difference in how hard your first second has to work.

The more your video looks like an ad, the faster it gets treated like one. Polish is a signal, and the signal says "skip me."

Why raw video builds trust faster

There is a second reason native video sells, and it runs deeper than the skip reflex. Rough edges read as honesty.

A glossy ad signals budget, agency, scripting, approval rounds. The viewer knows many people shaped every word. A person holding a phone, speaking plainly about something they know, signals the opposite: this is someone telling you what they actually think. Whether it is a founder, an employee or a customer, the lack of production makes the message feel less manufactured and more believable.

That is also why real customers on camera are so powerful. We break down the gap between customer content and paid influencers in real customers outsell influencers. And as more of the feed fills with synthetic, AI-generated imagery, the value of visibly real people keeps rising, a shift we cover in everything looks like AI now, that is your opening.

Authenticity is a strategy, not an accident

Here is the twist: "raw and authentic" does not mean careless. The best native-style videos are engineered to feel spontaneous. The hook is deliberate, the story is structured, the message is sharp, it is just dressed in the clothes of real, unpolished content instead of a glossy commercial.

• Hook in the first second, native content lives or dies on the opening frame.
• Talk like a human, not like a brand reading a script.
• Match the platform so it blends into the feed instead of interrupting it.
• Lead with value or story, and earn the pitch.
• Keep the craft invisible, structured underneath, effortless on the surface.

"Ugly" is a slightly misleading word. The best phone videos are not ugly. They are unpolished. There is a big difference. Bad audio, a dark room, a mumbled message and a lens smudge are not authentic, they are just bad. Viewers forgive a lack of gloss. They do not forgive being unable to hear you.

How to make a native-style video that sells

1. Start with the hook, not the brand. Open on the problem, the result or a surprising statement. Your logo can wait until the end, or be left out entirely.
2. Pick the right person. The founder, a technician, a real customer. Whoever the viewer will believe most on this topic. Rarely an actor.
3. Write bullet points, not a script. A word-for-word script gets read, and reading sounds like reading. Three or four points in the right order, delivered in the person's own words, sounds real.
4. Shoot vertical, on a phone, in a real place. The workshop, the job site, the kitchen, the car. The setting is part of the proof.
5. Fix the two things that matter. Clear audio with a simple clip-on microphone, and decent light from a window or a small panel. Everything else can stay rough.
6. Add captions. Plenty of people watch with the sound off. Captions keep them in the story.
7. End with one clear action. Book, buy, message, call. One ask, stated plainly.

The structure underneath should still follow a proven shape: hook, story, offer. The phone changes the costume, not the skeleton.

Hooks that work in native video

The opening line does most of the work, so it deserves most of your planning time. A few patterns that suit the phone format well:

• The mistake: "Most people clean their oven like this, and it is why it never looks clean."
• The result first: open on the finished job, then rewind to how it happened.
• The blunt opinion: something your industry quietly thinks and rarely says.
• The customer question: "Someone asked me yesterday whether..." followed by the answer.

Film three or four different hooks for the same video. It costs a few extra minutes and gives you genuinely different ads to test.

A worked example: the landscaper and the drone ad

Say you run a landscaping business in Newcastle and you spend $6,000 on a brand commercial. Drone flyovers of finished gardens, slow pans across stonework, soft music, a voiceover about "transforming outdoor spaces". It looks beautiful. You run it as a social ad and it gets skipped almost immediately. People watch a second or two, register "ad", and move on.

Now picture the alternative. You stand in a client's backyard (with their permission) holding your phone, and the first line is: "This backyard flooded every time it rained, and here is the ten-dollar mistake that caused it." You walk the viewer through the drainage fix in forty seconds, point at the problem, show the solution, and finish with "If your yard does this, send me a photo and I will tell you if it is the same problem."

No drone. No music bed. A useful story, told by someone who clearly knows what they are doing, in the visual language of the feed. You could film several variations of this in a single morning for almost nothing. This is a hypothetical, but it shows why the second approach tends to earn attention the first one never gets. And the drone footage? It still has a job on your website, where people have already chosen to look.

Not either/or, the right tool for the job

Polished brand films still have their place for trust and prestige. But for stopping the scroll and driving action in the feed, native-style video usually wins. The brands cleaning up right now know which tool to reach for, and how to make "authentic" look effortless while engineering every second of it.

A simple way to split it:

• Polished works best on your homepage, in a pitch, at an event, on a product page, in a brand story. Places where the viewer has chosen to pay attention and you need to signal quality. We cover how to make those count in the anatomy of a brand video that sells.
• Native works best in paid social, organic short-form, stories, and retargeting. Places where you are interrupting someone and need to earn the next three seconds.
• Both can come from one shoot. Plan a day with a proper crew for the hero piece, and build in time to capture phone-style clips, talking heads and customer moments alongside it.

The mistakes that make phone videos flop too

• Faking it badly. A scripted actor pretending to be a real customer is easy to spot and destroys trust. Use real people.
• Unusable audio. Wind noise, echo and traffic kill more native videos than any lack of polish.
• No hook. Raw does not excuse a slow start. "Hi guys, so today I wanted to talk about" is a skip.
• Burying the offer. Authentic content still needs to tell people what to do next.
• Making one and stopping. Native ads wear out fast. You need a steady supply, as we explain in your winning ad now dies in 14 days.

How to test raw against polished

Do not take anyone's word for it, including ours. Run your polished ad and two or three native-style versions side by side in the same campaign, with the same budget and the same offer. Judge on cost per result and on how long people watch, not on which one the team likes most in the review meeting. Give it enough spend to see a pattern before calling a winner.

Whatever wins, look at why. Was it the hook? The person? The setting? That lesson is worth more than any single ad, because it tells you what to make next.

Fit beats finish

We produce content built for how people actually watch, native-feeling video engineered to convert, plus the polished work when the moment calls for it. Our photo and video team at LIQID NOISE plans both from the same shoot day, so you get the hero film and a stack of feed-ready clips without paying twice. If your beautiful ads are getting skipped, the problem is not quality. It is fit. Let's make content the feed cannot ignore.

Frequently asked questions

Do phone videos perform better than professional ads on social media?

Often, yes, in the feed. Native-looking phone video blends in with the content people came to watch, so it is less likely to be skipped on sight. That gives your message a chance to land. Professional production still performs well where people have chosen to pay attention, like websites and presentations. The best approach is to test both with the same offer.

What equipment do I need to film ads on my phone?

Less than you think. A recent smartphone, a simple clip-on or wireless microphone for clear audio, and good light from a window or a small LED panel cover most needs. A basic phone tripod helps with steady talking-head shots. Prioritise sound and light first, because viewers forgive a lack of polish far more readily than muffled audio.

Should my ads look like user-generated content?

They should feel native to the platform, but never deceive. Real customers, real staff and real founders speaking in their own words are ideal. Hiring actors to pretend to be customers risks trust and may breach advertising rules in some places. Aim for the style of user-generated content with honest people behind it.

How long should a native-style video ad be?

As long as it holds attention and no longer. Many strong native ads run well under a minute, but the length matters less than the first seconds. If the hook is strong and every line earns the next, viewers will stay. Test a short cut and a longer cut of the same idea and let the results guide you.

So how do you retarget without being creepy? Segment people by how close they are to buying, show each group a different message that moves the story forward, cap how often anyone sees you, exclude people who've already bought, and prove the spend is actually creating sales with a holdout test. That's the whole system. The rest of this article is how to build it.

Most accounts get none of this right. They run one "all website visitors, last 180 days" audience with one ad on repeat and wonder why results slide and comments turn nasty.

Why retargeting goes wrong

Retargeting fails for a predictable reason: it's set up once and then left alone. Someone builds a single warm audience, drops in a product ad, and the platform does what platforms do. It finds the cheapest impressions available, which are the same few people, over and over.

From the advertiser's side, the numbers look fine for a while. Retargeting audiences convert well because they were already interested. But underneath, three things are happening:

• The same people are being hit far too often. What felt like a friendly reminder on day two feels like being followed by day ten.
• Everyone gets the same message. Someone who glanced at your homepage and someone who abandoned a full cart are treated identically, which means the message fits neither.
• The credit is inflated. A lot of the people who "converted from retargeting" would have come back anyway.

Fix those three and retargeting goes from irritating to genuinely useful, for you and for the people seeing it.

Segment by intent, not just visits

A homepage bouncer and a cart abandoner are not the same person. They deserve different messages, different offers, and different urgency. Map your audiences to where they actually are in the decision:

• Top: Reintroduce the brand and the core value. No hard sell.
• Middle: Handle objections, social proof, comparisons, guarantees.
• Bottom: Remove friction. A reminder, an incentive, a one-tap path back.

In practice, you'd build audiences like these:

• Engagers: people who watched a good chunk of a video or interacted with your page, but never visited the site. Top of funnel.
• Browsers: site visitors who viewed a page or two and left. Top to middle.
• Considerers: visitors who viewed specific product or service pages, pricing, or spent real time on the site. Middle.
• Hand-raisers: people who added to cart, started checkout or began a form. Bottom.
• Buyers: excluded from the sales sequence, and moved into a separate audience for repeat purchase or referral offers where that makes sense.

Each audience should exclude the ones below it, so people move cleanly down the funnel instead of seeing three sequences at once.

Frequency caps aren't a limitation. They're respect, and respect converts.

Sequence the story

Great retargeting feels like a conversation, not a loop. Each impression should advance the narrative instead of repeating it. We build sequenced funnels where the creative evolves as the relationship warms, so by the time someone buys, it feels like their idea.

A simple sequence for a considered purchase might run like this:

1. Days one to three: the "why us." A short video or carousel that explains what makes you different. No discount, no pressure.
2. Days four to seven: the proof. Customer stories, reviews, before-and-after results, behind-the-scenes of how you work.
3. Days eight to fourteen: the objection handler. Answer the question that stops people buying. Price, timing, risk, "will this work for me." A guarantee or a comparison goes here.
4. Final days: the nudge. A clear reminder with a simple path back. If you use an incentive, it lives here and only here.

The timing will change depending on how long your buyers take to decide. An impulse product might compress the whole thing into a few days. A high-ticket service might stretch it across a month. The principle stays the same: never show someone the same thing twice when you could show them the next thing.

Creative fatigue is real, and it hits retargeting hardest because the audience is small. If you want to understand how quickly winning ads burn out now, read why your winning ad dies faster than it used to.

Cap the frequency before the algorithm decides for you

Left alone, ad platforms can serve the same person the same creative again and again in a week, because that is often the cheapest inventory they have. That is not efficiency. That is burning goodwill you paid to earn.

• Set a real ceiling: Three to five impressions per person per week for a considered purchase. Higher for impulse buys, far lower for anything expensive.
• Set an expiry: Every audience needs one. Someone who visited ninety days ago and never came back is not warm, they are cold traffic you are overpaying to reach.
• Suppress your buyers: Nothing torches trust faster than advertising a product to the person who bought it yesterday. Exclusion lists are the highest-return five minutes in the whole account.

Suppression depends on the platform actually knowing who bought. If your purchase tracking is patchy, buyers slip through and keep seeing ads. That's one of many reasons to fix your conversion data, which we cover in the boring number quietly deciding your ad results. Uploading your customer list directly as an exclusion audience is a useful backstop.

The goal is not to be seen more often. It is to be seen at the moment the decision is actually being made.

A worked example: the online homewares store

Here's a hypothetical. Say you run an Australian online homewares store spending $3,000 a month on Meta ads, with about a third of that going to one retargeting campaign aimed at everyone who visited in the last 180 days. It's one carousel of best-sellers, running for months. Comments are starting to include "stop following me."

Rebuilt properly, the setup could look like this:

• Browsers from the last 14 days: a short brand video on how the pieces are sourced and made, plus a carousel of styled rooms. Capped at a few impressions a week.
• Product viewers from the last 14 days: reviews and customer photos of the specific categories they viewed, then a piece answering the big objections: delivery times, returns, how colours look in real homes.
• Cart abandoners from the last 7 days: a clean reminder of the item with a one-tap path back to the cart and a clear note on free returns.
• Buyers: excluded from all of the above. After a sensible gap, a separate audience sees complementary products.
• Holdout: a slice of each audience never sees ads, so you can measure what the campaign is truly adding.

Same budget. Fewer wasted impressions, messages that fit where each person is, and an honest number at the end of the month telling you whether it's working.

Measure incrementality, not credit

Retargeting always looks brilliant in a last-click report, because it is standing closest to the finish line taking credit for a race someone else ran. Run a holdout instead. Keep a slice of the audience unexposed and compare conversion rates. If the exposed group does not clearly outperform, you are paying to reach people who were coming back anyway.

How to run one without a data science team:

1. Use the platform's built-in lift or experiment tools where they exist. They handle the random split for you.
2. If you can't, split manually. Randomly divide a customer or visitor list and only target one half. Compare purchase rates across both halves over the same period.
3. Run it long enough. At least a full buying cycle. A few days of data will mislead you.
4. Look at the difference, not the total. The value of retargeting is the gap between the exposed and unexposed groups, not how many sales the exposed group made.

This is the same discipline that uncovers waste across the whole account. If you suspect your spend is leaking elsewhere too, our ad spend audit walks through where to look.

Common retargeting mistakes

• One audience for everyone. The message fits nobody.
• No expiry on audiences. You're paying warm prices for cold people.
• Discounting too early. Offer a discount on the first retargeting ad and you train visitors to leave and wait for one.
• Forgetting buyers. Advertising the thing they just bought makes you look careless.
• Same creative for months. Small audiences see ads more often, so they tire of them faster.
• Trusting last-click. It flatters retargeting every time.

What to fix first this week

1. Check your frequency. Open your retargeting campaigns and look at average frequency over the last seven days. Anything well above five for a considered purchase is a warning sign.
2. Add buyer exclusions. Exclude purchasers from every sales campaign, using both your tracking and an uploaded customer list.
3. Shorten your windows. Replace one long catch-all audience with at least two: recent visitors and high-intent visitors.
4. Write the objection ad. Ask your sales team or inbox what stops people buying, and make one ad that answers it directly.
5. Set up a holdout. Even a rough one tells you more than a month of last-click reports.

Be the brand people are glad to see again

Retargeting done badly teaches people to resent you. Retargeting done well feels like a helpful follow-up from a business that remembered what you were interested in and has something useful to add. The difference is not budget. It's segmentation, sequencing, restraint and honest measurement.

That's how our advertising team builds retargeting: audiences mapped to intent, creative that moves the story forward, frequency caps and exclusions set on day one, and holdout tests that prove the spend is earning its keep. If your current setup is one ad chasing everyone for six months, there's money being wasted and goodwill being burned. Let's fix both.

Frequently asked questions

How often should a retargeting ad be shown to the same person?

For a considered purchase, around three to five impressions per person per week is a sensible ceiling. Impulse products can tolerate more, while expensive or sensitive purchases need far less. Rotate creative within that cap so people see a progression of messages rather than the same ad. If comments or frequency metrics show irritation, lower the cap and refresh the creative.

How long should retargeting audiences last?

Match the audience window to how long your customers take to decide. Cart abandoners are warmest within a week, general site visitors within a few weeks, and considered services may justify a longer window. Every audience needs an expiry. Someone who visited months ago and never returned behaves like cold traffic, so paying retargeting prices to reach them is usually wasteful.

How do I know if retargeting is actually working?

Run a holdout test. Keep a random portion of your retargeting audience from seeing ads and compare their conversion rate with the exposed group over a full buying cycle. If the exposed group converts meaningfully better, retargeting is adding sales. If the rates are similar, you are mostly paying to reach people who would have returned anyway, regardless of what last-click reports suggest.

Should I offer a discount in retargeting ads?

Only at the end of a sequence, and only if your margins and brand allow it. Leading with a discount teaches visitors to leave your site and wait for one, which erodes full-price sales. Start with your value, proof and answers to objections. Reserve any incentive for people who have shown strong intent, such as cart abandoners, and test whether it genuinely lifts conversions.

So how do you write a business book people actually finish? You make one clear promise, build every chapter around a single idea, carry each idea on a real story, write in your own speaking voice, and end each chapter with a reason to keep going. That is the whole method. The rest of this article shows you how to do each part.

Boring is a choice, not a genre

Nobody decides business writing has to be dull, it just drifts that way when authors try to sound impressive instead of trying to be read. They pile on jargon, hedge every point, and bury insight under throat-clearing. But the books that actually change careers and fill pipelines read almost like a conversation: clear, human, and impossible to put down.

Watch what happens when a smart founder sits down to write. On a sales call they are sharp, funny and specific. They tell stories, they use plain words, they get to the point because the prospect is watching the clock. Then they open a blank document titled "Chapter One" and something changes. They start writing like they are submitting a thesis. "In the contemporary business environment, organisations face unprecedented challenges." Nobody talks like that. Nobody wants to read it either.

The cause is fear. Fear of being judged by peers, fear of being too simple, fear of leaving something out. So the author covers every base, qualifies every claim and adds every idea they have ever had. What comes out is defensible and unreadable.

A book nobody finishes is a brochure nobody keeps. Getting read is not a bonus, it is the entire job.

Why an unfinished book costs you more than sales

Most authors of business books are not trying to earn a living from royalties. They want the book to open doors: speaking gigs, consulting work, higher fees, a reputation as the person who owns a topic. We make the full case for that in write the book that closes deals while you sleep.

Every one of those outcomes depends on the reader getting to the good part. Your method, your framework, your offer at the back of the book. If they quit in chapter two, they never see the thinking that would have made them call you. Worse, they remember you as the author of the book they gave up on. That is not neutral. That is a small, lasting mark against your name.

A finished book works the other way. The reader has spent hours with your ideas in your voice. By the last page they feel like they know you. That is a level of trust no ad, post or website can create, and it is exactly why a readable book is such a powerful sales asset.

What keeps readers turning pages

• Story. Ideas wrapped in real stories stick, abstract lectures slide off.
• A clear promise up front, and a structure that delivers on it step by step.
• One idea per chapter, made vivid, instead of ten ideas made muddy.
• Voice. It should sound like a smart human talking, not a committee writing.
• Momentum, every chapter ending in a way that makes stopping feel impossible.

Each of these is a craft skill, not a talent. Let's go through them properly.

Story

People remember what happened to someone far longer than they remember a principle. So every major idea needs a vehicle: a client situation (anonymised if needed), a mistake you made, a moment you figured something out. The best structure is simple. What was the situation, what went wrong or what was at stake, what changed, and what that teaches. Tell the story first, then name the lesson. Most authors do it backwards and lose the reader before the story arrives.

A clear promise

Within the first few pages the reader should know exactly what they will be able to do, understand or decide by the end. Not "this book explores leadership". Something like "by the end of this book you will be able to run a team meeting that people leave with decisions, not more meetings". A clear promise gives the reader a reason to keep going and gives you a test for every chapter: does this move them toward the promise? If not, cut it.

One idea per chapter

A chapter with one idea can be explained, illustrated, stress-tested and applied. A chapter with five ideas can only be listed. If you find a chapter doing three jobs, split it or cut two of them. Readers finish books that feel like a series of clear steps, and abandon books that feel like a pile.

Voice

The quickest fix for stiff prose is to talk it out. Record yourself explaining the chapter to a client, transcribe it, and edit that. You will keep your natural rhythm, your turns of phrase and your examples. Then run a jargon sweep: any word you would not say across a table, replace.

Momentum

End chapters on a question, a tension or a preview of the next problem. "That solves the pricing issue. But it creates a new one, and it is the one that sinks most businesses in their second year." Television writers do this all the time. Plenty of business authors never do, and it is a free advantage.

How to structure a book that holds attention

Structure is where readability is won or lost, long before a sentence is written. Here is a simple build order that works for most expert-led business books.

1. Write the promise in one sentence. What the reader will be able to do by the end.
2. List the steps between where they are and that promise. Usually seven to twelve. Each becomes a chapter.
3. Assign one story to each step. If a step has no story, you either need to find one or question whether it belongs.
4. Write a one-line takeaway for each chapter. If you cannot, the chapter is not clear yet.
5. Plan the chapter endings. What question or tension carries the reader into the next one?
6. Only then start drafting. With this map, drafting becomes filling in, not figuring out.

This outline also makes the book far faster to write. Busy founders who do it well can go from idea to finished manuscript in months, not years, which is the approach we lay out in the 90-day book for busy founders.

A worked example: fixing a chapter that nobody finishes

Say you are a financial adviser in Melbourne writing a book for small business owners about paying themselves properly. Your first draft of chapter three opens with four paragraphs on the history of superannuation, then introduces three different frameworks for owner salaries, then lists tax considerations, then ends with "in the next chapter we will look at cash flow".

Every piece of it is accurate. None of it is readable.

Now rebuild it. Open with a story: a café owner (a composite, details changed) who worked seventy-hour weeks for four years and paid herself less than her junior barista, until the morning she could not cover her own rent. Name the idea in one sentence: owners who pay themselves last end up running a business that cannot afford them. Give one framework, not three, and walk the café owner through it with simple numbers. Add a short "do this tonight" exercise. End with tension: "Once she paid herself properly, the business looked broke on paper. It was not. But she needed to see cash flow differently, which is chapter four."

Same expertise. Same author. One version gets skipped. The other gets finished, underlined and handed to a friend. That is a hypothetical, but it is the exact edit that separates the two kinds of book.

Common mistakes that make business books unreadable

• The long wind-up. Forewords, acknowledgements, a preface and a twenty-page introduction before anything useful. Get to the first insight fast.
• Proving credentials instead of helping. Your expertise shows through the quality of the advice, not a chapter about your career.
• Hedging everything. "In some cases, it may be worth considering." Say what you think. Add caveats once, where they genuinely matter.
• Walls of text. Short paragraphs, clear subheadings and the occasional list give the eye a place to rest.
• Everything you know. The book is not your complete knowledge. It is the path to one promise. The rest can be your next book.
• Editing for grammar, not for boredom. Proofreading catches typos. It does not catch the three pages where the reader quietly stopped caring.

How to test readability before you publish

You do not need to guess. Before the book goes to print, give chapters to a handful of people who match your ideal reader, not your peers or your family. Ask them to mark the exact point where their attention drifted and the point where they wanted to share something. Those marks are gold. Fix the drift points. Protect the share points.

Read the whole manuscript aloud, too. Anywhere you stumble, run out of breath or feel bored hearing your own words, the reader will feel it twice as strongly. And once it is ready, the launch matters as much as the writing, which we cover in your book launch is where most authors die.

Readable is what makes it work

This is not about "dumbing down." It is about respecting the reader enough to be clear, engaging, and useful on every page. Because a book that gets finished is a book that gets recommended, quoted, and acted on, and that is the book that actually turns readers into clients and authority into revenue.

We help experts turn deep knowledge into books that are genuinely gripping to read, clear, story-driven, and built to move the reader to action. Through our books and publishing service, LIQID NOISE works with you from promise and outline through to drafting, editing and launch. If you want a book that works instead of one that gathers dust, it has to be read first. Let's write one nobody can put down.

Frequently asked questions

How long should a business book be?

Long enough to deliver on its promise and no longer. Many strong business books are shorter than authors expect, because they cover one clear path instead of everything the author knows. Build your outline around the steps to your promise, give each a chapter, and let the length follow. If a chapter exists only to make the book thicker, readers will notice and stop.

How do I make my non-fiction writing more engaging?

Lead with stories before principles, keep one idea per chapter, and write the way you speak to a client. Short paragraphs and clear subheadings help. End each chapter with a question or tension that pulls the reader forward. Reading your draft aloud is the fastest test, because anywhere you get bored or stumble, a reader will too.

Can I use client stories in a business book?

Yes, and you should, but protect your clients. Get permission for anything identifiable, or change names, industries and details so nobody can be recognised. Composite stories built from several real situations are common and useful, as long as you do not present them as one specific real person. The lesson carries the value, not the identity of the client.

Why do readers stop reading business books?

Usually because the book takes too long to get useful, crams too many ideas into each chapter, or reads like a report instead of a person talking. Jargon, hedging and long introductions all drain momentum. Readers keep going when they feel progress toward a clear promise, meet vivid stories and hit chapter endings that make them curious about what comes next.

So how do you name a brand people actually remember? Pick a name that's easy to say, easy to spell after hearing it once, clearly different from your competitors and big enough to grow into. Then test it out loud with real people and lock down the domain, handles and trademark before you announce it. That's the whole game. The hard part is having the discipline to follow it when you've fallen for a name that fails.

Because that's what usually happens. A founder gets attached to a clever pun, an in-joke or a word from a language nobody in their market speaks, and spends the next five years spelling it down the phone. Let's make sure that isn't you.

The tests every name must pass

• Sayable: If people hesitate to pronounce it, they won't recommend it.
• Spellable: Heard once, typed correctly. Domains and search depend on it.
• Distinct: It should stand apart in a crowded category, not blend in.
• Expandable: Don't box yourself into a single product or city you'll outgrow.

Each test protects a different part of your growth. Sayable protects word of mouth, the cheapest marketing channel you'll ever have. Spellable protects your search traffic and your email deliverability. Distinct protects your memorability, and your legal position when it comes to trademarks. Expandable protects the future version of your business that you haven't built yet.

Notice what's not on the list: clever. Clever is nice. It is not a requirement. A plain name that passes all four tests will outperform a brilliant name that fails one.

A name doesn't have to describe what you do. It has to be impossible to forget.

The five kinds of brand name

Almost every brand name falls into one of five types. Knowing the trade-offs helps you choose on purpose instead of by gut feel.

• Descriptive. Says exactly what you do, like "Brisbane Roof Repairs." Instantly clear and handy for local search. But it's hard to trademark, easy to confuse with competitors and hard to stretch if you expand.
• Invented. A made-up word. Highly distinctive, easier to protect and easier to find a domain for. But it starts with zero meaning, so you have to pay to teach people what it stands for.
• Evocative. A real word used in a new context that hints at a feeling or benefit. Memorable and flexible, but popular evocative words are often already taken in your category.
• Founder or personal. Named after a person. Warm and trustworthy, great for service businesses. Harder to sell or scale beyond the founder.
• Acronym. Initials. Compact, but usually forgettable unless you already have a big brand behind it. Avoid starting with one.

There's no universally right type. A local tradie may win with a descriptive name. A product brand planning to go global usually needs something more distinctive. The right type depends on your positioning, which is why we argue positioning comes before the logo, and before the name too.

Meaning is built, not found

Some of the strongest brands started as invented or abstract words. They earned meaning through consistency. Your name is an empty vessel on day one, what fills it is everything you do afterward.

That's liberating. It means you can stop hunting for the "perfect" word that explains everything. No name can carry your entire value proposition. The name's job is to be a clean, memorable hook. Your message, your visuals, your service and your reputation fill it with meaning. If your name is doing all the explaining, your messaging probably isn't doing enough, which we unpack in your message is the problem.

How to generate a shortlist that doesn't suck

Good names rarely arrive in a flash of inspiration. They come from volume, then ruthless cutting. Here's the process:

1. Write a one-page brief. Who you serve, what you stand for, how you want to feel, who your competitors are and what they're called. Naming without a brief is guessing.
2. List your raw material. Words about the outcome you deliver, the feeling you create, your origin story, your values and the imagery of your industry.
3. Generate in bulk. Aim for a long list, at least a hundred options. Combine words, shorten them, borrow from other fields, change a letter, try metaphors. AI tools are genuinely useful here as a brainstorming partner, as long as you don't let them pick for you.
4. Kill fast. Cut anything that fails the four core tests. Be merciless.
5. Check availability early. Do quick domain, social and trademark searches on the survivors before anyone gets attached.
6. Shortlist three to five. Mock each one up on a logo, a website header and a business card. Names feel different once they look real.

Test the name in the real world before you commit

Most founders fall in love with a name inside a document. That is the worst possible place to judge one. A name only proves itself out loud, in a noisy room, spoken by someone who has never seen it written down.

• The phone test: Say it to five people and ask them to type it into a browser. If more than one gets it wrong, the name is leaking traffic before you have even launched.
• The trademark test: Search the register in every market you intend to sell in. Falling in love first and checking later is how brands end up rebranding in year three.
• The shout test: It has to survive a bad phone line, a busy trade show floor and a podcast host reading it cold.

Then check what the internet already thinks. Search the name, look at what comes back, and work out whether you are launching into an open lane or an existing conversation you cannot win. That now includes asking the AI assistants directly, because a growing share of buyers meet a brand through an answer rather than a search result.

A few more checks worth doing before you commit:

• The meaning check. Look the name up in other languages spoken in your markets, and in slang dictionaries. Embarrassing surprises are common.
• The confusion check. Is there a similar-sounding name in your category or a neighbouring one? Confusion costs you customers and can bring a legal letter.
• The entity check. A name that's already a common word, place or famous person makes it harder for search engines and AI tools to know you're a distinct business. That matters more every year, as we explain in become an entity.

A worked example: naming a new bakery

Say you're opening a sourdough bakery in Geelong, with plans to add wholesale and a second shop in a few years.

Your first idea is "Geelong Sourdough Co." It's clear, but it fails the expandable test the moment you open elsewhere or add pastries, and it's barely distinct from other local bakeries. Your second idea is a clever French phrase you love. It fails the phone test: three of five friends can't spell it and two can't pronounce it.

So you go back to the brief. The bakery's story is about slow bread, early mornings and a family starter kept alive for years. You generate a long list around those ideas and land on a short, evocative English word linked to that early morning ritual. Everyone spells it after hearing it once. It works on a delivery van and a wholesale invoice. The domain and handles are available, and a trademark search in your classes comes back clear. That's the name. Not the cleverest option, the strongest one.

And if you already have a name that fails these tests? Don't panic and rebrand overnight. Weigh the daily friction it causes, the misspelled emails, the confused referrals, the search results full of someone else, against the recognition you've already earned. Sometimes a small fix, like a clearer descriptor or a simplified spelling, solves most of the problem. Sometimes a full rename is the right call, and it's far cheaper done now than in five years.

Mistakes that cost founders years

• Naming by committee. Asking everyone for a vote produces the least offensive, least memorable name. Get feedback on the tests, not opinions on taste.
• Chasing the exact-match domain. A great name with a sensible domain variation beats a mediocre name with a perfect domain.
• Trend spelling. Dropped vowels and swapped letters look dated fast and fail the phone test every time.
• Too literal. A name that describes one product becomes a problem the day you launch a second.
• Skipping legal checks. Rebranding after launch means reprinting, rebuilding and re-earning recognition.

Lock it down on day one

Secure the domain, the handles and the trademark before you tell anyone outside the room. Names leak. The cost of taking all three up front is trivial next to the cost of rebuilding recognition you have already paid for.

Your day one checklist:

1. Register the main domain and any obvious misspellings or variations.
2. Claim the handle on every social platform you might use, even the ones you won't use yet.
3. File a trademark application in the relevant classes, and get advice from a trademark professional if you're unsure.
4. Register the business name where your jurisdiction requires it.
5. Set up a Google Business Profile if you serve local customers.

A name with room to grow

A great name won't build your business. A bad one will quietly slow it down every single day. Get it right early and it becomes an asset that gathers value with every sale, review and conversation, which matters more than ever now that brand is the last moat.

We help founders choose a name with room to grow, then build the world that gives it weight. Our brand consulting covers positioning, naming, messaging and the identity that brings it all to life. If you're naming something new or wondering whether your current name is holding you back, let's talk.

Frequently asked questions

How do I check if a business name is already trademarked?

Search the official trademark register in each country you plan to trade in. In Australia, that's IP Australia's online trade mark search. Look for identical and similar names in the classes that cover your goods or services. A clear search is a good sign but not a guarantee, so get professional advice before investing heavily in a name.

Should my brand name describe what I do?

Not necessarily. Descriptive names are clear and can help local search, but they're harder to trademark, easier to confuse with competitors and limiting if you expand. Distinctive invented or evocative names are easier to protect and remember, though they need marketing to build meaning. Choose based on your positioning and how far you plan to grow.

What if the .com domain for my brand name is taken?

It's often not a dealbreaker. Consider your local domain, such as .com.au in Australia, or a simple, logical variation like adding a word that describes what you do. Avoid awkward spellings or hyphens. Check that the owner of the taken domain isn't a competitor or something you'd hate to be confused with.

Can I use AI to name my business?

AI is a useful brainstorming tool for generating long lists of name ideas quickly. It shouldn't make the final decision. Every suggestion still needs to pass the sayable, spellable, distinct and expandable tests, plus real availability and trademark checks. Treat AI output as raw material, then test your shortlist with real people.

Get the structure right and a landing page becomes the closest thing in business to a money-printing machine, put a dollar of traffic in one end, get more than a dollar of customers out the other. Get it wrong and you are lighting your ad budget on fire.

So what makes a landing page convert? A headline that makes one specific promise, proof placed exactly where doubt shows up, an offer worth more than the price, and a single call to action with every other exit sealed. Everything else is decoration. Below is the full anatomy, the order it goes in, and how to test it once it's live.

The elements that do the heavy lifting

• The headline, a promise so specific and desirable the right person cannot look away. This is most of the game.
• The subhead, how you deliver that promise, in one believable sentence.
• The proof, testimonials, results, guarantees, and numbers placed exactly where scepticism appears.
• The offer, framed so the value dwarfs the price and saying no feels like the risky choice.
• The single call to action, repeated down the page, always the same next step.

Every extra choice you add to a landing page is another exit you carved for the customer you were trying to keep.

Let's take them one at a time, because each has a job, and most pages fail at least one of them.

The headline: match the click

The visitor arrived with a specific desire, usually shaped by the ad or search they just clicked. Your headline's first job is to confirm, instantly, that they landed in the right place. If the ad said "Emergency plumber in 60 minutes" and the page says "Welcome to Smith Plumbing Services", you've broken the thread. Repeat the promise that earned the click, then sharpen it.

The subhead: make it believable

The headline makes the promise. The subhead explains how, in one line that makes the promise feel possible. "Licensed local plumbers on call 24/7 across the inner west, with upfront fixed pricing." Specific, concrete, credible.

The proof: put it where the doubt lives

Proof isn't a testimonials section dumped at the bottom. Think about the objections running through the visitor's head as they scroll: will this work for someone like me, is this price fair, can I trust these people? Place a specific piece of proof right next to each one. A review about price next to the price. A before-and-after next to the claim. A guarantee next to the button.

The offer: make no feel risky

The page can only convert as well as the offer lets it. If what you're offering is the same as every competitor, no layout will save you. We break down how to build an offer worth landing on in the Godfather Offer.

The call to action: one door, many handles

Choose one action. Book a call, get a quote, start a trial, buy now. Then repeat that same button through the page, so whenever the visitor is convinced, the next step is right there. Write the button as an outcome, "Get my fixed quote", not a chore like "Submit".

Why most landing pages leak

They hedge. They add a nav bar "just in case." They link out to the blog, the about page, the social accounts. Every one of those links is a door the visitor can leave through before they buy. A page that converts is ruthless about doors: it seals every one except the one marked "yes."

It also respects the order of persuasion. You do not ask for the sale before you have earned it. You lead with the outcome, agitate the problem, prove you can solve it, then, and only then, make the ask. The page carries the visitor over their objections one at a time until saying yes is the natural next move.

Other common leaks:

• Sending ad traffic to the homepage. The homepage is built for everyone, which means it's built for no one in particular. Paid traffic deserves a page built for the exact promise in the ad.
• Forms that ask for everything. Every extra field is another reason to quit. Ask only what you need to take the next step.
• Slow load times. A visitor who leaves before the page appears never sees your brilliant headline. We cover the damage in why a slow website is a silent business killer.
• A desktop-first design. A big share of ad clicks happen on phones. If the button is below three screens of scrolling on mobile, you've hidden the one door you wanted open.
• Vague claims. "Quality service" and "trusted experts" are invisible. Numbers, timeframes and specific outcomes are what the eye catches.

The order of persuasion, section by section

Here's a proven structure you can build almost any high-converting landing page around. Adapt it, but keep the order.

1. Hero: headline that matches the ad, subhead that makes it believable, the call to action button, and one strong trust signal such as a review rating or a guarantee.
2. The problem: describe the visitor's situation in their own words, so they feel understood before they feel sold to.
3. The cost of doing nothing: what staying stuck is costing them in money, time or stress.
4. The solution: how your offer fixes it, framed around outcomes, not features.
5. Proof: testimonials, results and examples, each chosen to answer a specific objection.
6. The offer: exactly what they get, what it costs and why it's worth more than the price.
7. Risk reversal: the guarantee, stated plainly.
8. FAQs: the last few objections, answered honestly.
9. Final call to action: the same button, with a line reminding them what they'll get.

A worked example: the solar installer's two pages

Here's a hypothetical. Say you run a solar installation business in Newcastle, spending $5,000 a month on Google Ads for searches like "solar panels quote Newcastle". Right now every click lands on your homepage, which has a slider, a menu with twelve links, a company history and a contact form at the bottom.

Visitors arrive wanting a quote. They're greeted by a carousel about your values. Some click into "About". Some click "Products" and get lost comparing panel brands. Some find the form eventually. Many just hit back and click the next ad.

Now build a dedicated page. The headline mirrors the search: "Get a fixed solar quote for your Newcastle home within 24 hours." The subhead explains who installs it and how long it takes. Reviews from local homeowners sit right under the hero. A short section explains how the quote works. The form asks for postcode, roof type and phone number, nothing else. There's no menu. Every button says "Get my free quote".

Same ads. Same budget. Same traffic. But now every visitor faces one clear path, and far more of the clicks you're already paying for turn into quote requests. That's the whole trick. You didn't buy more traffic. You stopped wasting what you had. If your ads are the bigger leak, pair this with the one thing that turns Google Ads from a money pit into a machine.

How to test and improve a landing page

A landing page is never finished. It's a machine you tune. Here's how to do it without guessing.

• Measure the right number. Conversion rate is visitors who took the action divided by total visitors. Track it per traffic source, because ad visitors and email visitors behave differently.
• Test big things first. The headline and the offer move results far more than button colours. Change the promise before you change the font.
• One change at a time. If you change five things and conversions rise, you don't know which one worked.
• Wait for enough data. A handful of conversions isn't a result. Let a test run until the difference is clear, not until you get bored.
• Watch real sessions. Session recordings show where people stop scrolling, what they tap that isn't a link, and where they leave. It's the fastest way to spot confusion.
• Check lead quality, not just quantity. A page that doubles enquiries but halves the close rate hasn't won anything.

If you only do one thing this week, do this: open your top ad on your phone, tap it, and look at where it lands with fresh eyes. Does the headline repeat the promise you just clicked on? Can you see the button without scrolling? Could you leave through a menu link before deciding? Every "no" is a leak, and each one is cheaper to fix than the traffic it's wasting.

Then plan for more than one page. Different audiences want different promises. A first-home buyer and a property investor might both need a mortgage broker, but they respond to completely different headlines and proof. One page per audience, per core offer, is how the best advertisers keep the message tight.

The compounding advantage

Here is what makes landing pages so powerful: once you have one that converts, you can pour traffic into it with total confidence. Every new channel, every campaign, every dollar of ad spend now has a reliable machine to run through. You stop guessing and start scaling.

And the gains stack. A better landing page lifts the return on every channel pointed at it. Better returns mean you can afford to bid more for traffic than competitors whose pages leak. More traffic means faster testing, which means a better page. That flywheel is why the businesses with the best landing pages so often end up owning their ad auctions.

Your website as a whole should work the same way, a point we push hard in your website is your best salesperson. The landing page is where that salesperson is at its sharpest: one prospect, one conversation, one close.

Stop hedging, start converting

That is what we build, not pretty pages, but engineered conversion assets tested against the only metric that matters: how many strangers became customers. If you are spending on traffic and sending it to a page that hedges, you are leaving most of your money on the table. Let's fix that.

Our web development team designs, writes and builds landing pages as one job, because copy, structure and speed only work when they're planned together.

Frequently asked questions

What is the difference between a landing page and a homepage?

A homepage introduces your whole business to every kind of visitor, so it offers many paths. A landing page is built for one audience, one promise and one action, usually for visitors arriving from a specific ad, email or campaign. Removing navigation and distractions is what lets a landing page convert at a much higher rate than a homepage.

Should a landing page have a navigation menu?

Usually not. Every menu link is an exit that takes visitors away from the one action you want. For paid traffic especially, remove the main navigation and keep only the elements that help the visitor decide: proof, the offer, FAQs and the call to action. A small logo and essential legal links in the footer are enough.

How long should a landing page be?

As long as it takes to answer the visitor's objections, and no longer. A simple, low-cost offer to a warm audience can convert with a short page. A bigger decision, a higher price or a cold audience usually needs more proof and explanation. Put the call to action near the top either way, then repeat it down the page.

What is a good landing page conversion rate?

It varies widely by industry, offer, price and traffic source, so benchmarks from other businesses are only a rough guide. The more useful question is whether your rate is improving and whether the leads are turning into customers. Track your own baseline per traffic source, then test headlines and offers to lift it.

Here is the short version. Mixing blends your individual recorded tracks into one balanced stereo song. Mastering takes that finished stereo song and prepares it to be released, so it sounds consistent, competitive and correct on every system. Mixing happens first. Mastering happens last. One cannot fix what the other got wrong.

That last line is where most independent artists burn money. They pay for a cheap mix, hope mastering will rescue it, and then wonder why the "professional master" still sounds muddy. Let's break down exactly what you're paying for at each stage, what to ask for, and how to tell if you got your money's worth.

Mixing: balancing the parts

Mixing is the art of taking every individual track, vocals, drums, bass, synths, and blending them into one cohesive whole. It's where the engineer carves space so nothing fights, sets levels, places elements in the stereo field, and shapes the emotional dynamics of the song.

In practice, a mix engineer is making hundreds of small decisions. Which instrument owns the low end, the kick or the bass? How far forward does the vocal sit? Should the guitars live hard left and right, or tucked in closer? When the chorus lands, what gets louder, brighter or wider so the listener feels the lift?

The main tools of mixing are simple to name and hard to master:

• Levels (volume balance). The foundation. Get this wrong and no plugin will save it.
• EQ (equalisation). Cutting and boosting frequencies so each instrument has its own lane instead of piling on top of the others.
• Compression. Controlling the gap between the quiet and loud moments of a performance so it sits steady in the song.
• Panning. Placing sounds left, right or centre to build width and separation.
• Space and effects. Reverb, delay and saturation that create depth, atmosphere and character.
• Automation. Changing any of the above over time, so the vocal rides up on a key line or the drums open up in the bridge.

A great mix is invisible. You don't notice it, you just feel the song the way it was meant to feel.

Mastering: the final polish

Mastering takes that finished stereo mix and prepares it for the world. It's the last quality check, tightening the overall tone, maximising loudness without crushing dynamics, and ensuring the track translates everywhere from earbuds to club systems to phone speakers.

The critical difference: a mastering engineer usually works on one stereo file, not dozens of separate tracks. They can't turn the snare down or pull the vocal forward. They can only adjust the whole song at once. That's why mastering is subtle by nature. Small moves in overall tone, a touch of compression, careful limiting to set the final level.

Mastering also handles the unglamorous but essential release work:

• Consistency across a project. On an EP or album, every track should feel like it belongs to the same record, with matched tone and level so nobody reaches for the volume knob between songs.
• Sequencing and spacing. The order of songs and the gaps between them.
• Delivery formats. The right file types for streaming, download, vinyl or CD, each with its own technical requirements.
• Metadata. Details like ISRC codes embedded so your plays and royalties can be tracked to you.
• A fresh set of ears. Often the most valuable part. A mastering engineer hears your song for the first time, in a treated room, with none of the emotional baggage you've built up over months of listening.

Why mastering can't fix a bad mix

This is the myth that costs artists the most. Mastering is not a repair shop. Because the engineer is working on the whole song as one file, any fix they make affects everything at once.

Say the vocal is buried. The mastering engineer boosts the frequencies where the voice lives. Now the vocal is a little clearer, but so are the guitars, the synth pads and the cymbals that share those frequencies. The song gets harsher and the vocal is still buried relative to everything else. The real fix was a level change in the mix that would have taken thirty seconds.

The same goes for boomy bass, a kick drum that disappears on small speakers, or a reverb tail that turns the chorus into soup. These are mix problems. A good mastering engineer will tell you so and send it back. A cheap one will just make it louder, which makes every problem louder too.

So the rule is simple. Put your budget into the mix first. A great mix with a decent master will beat a mediocre mix with a world-class master every time.

Why you need both

Mixing makes the song work. Mastering makes it compete. Skip either and listeners feel it even if they can't name it.

Skip the mix, or do it badly, and the song never hangs together. The parts fight. The hook doesn't land. Skip mastering and the song might sound great in your studio but thin and quiet next to everything else on a playlist, with tonal problems that only show up in the car or on a phone. Listeners don't sit there thinking "the low mids are cluttered." They just skip. If you want the full picture of what separates a finished release from a demo, we broke it down in what actually makes a track radio-ready.

The loudness trap most artists fall into

For years, the goal was simple: be louder than everyone else. Masters got pushed so hard that the life got squeezed out of them. Drums lost their punch, choruses stopped lifting, and everything turned into one flat wall of sound.

Streaming changed the rules. The major platforms apply loudness normalisation (Spotify turns louder masters down to around -14 LUFS, for example), which means they turn loud tracks down to a common playback level so listeners aren't constantly adjusting volume. Push your master too hard and the platform simply turns it down. You keep all the damage from over-compression and lose the advantage you were chasing.

The smarter goal now is a master that sounds full and punchy at a normal level, with enough dynamic range left that the song still breathes. Ask your mastering engineer how they approach loudness for streaming. If the only answer is "we make it as loud as possible," that's your cue to keep looking.

What you're actually paying for

Mixing and mastering rates vary wildly, from bargain online services to top-tier engineers. Instead of fixating on price, look at what the price includes. Here's what separates a good deal from a false economy:

1. The ears and the room. You're paying for years of trained listening and a properly treated space with accurate monitoring. That's the part you can't buy as a plugin.
2. Time per song. A proper mix takes hours of focused work. If someone is turning around full mixes in minutes, you are getting a preset, not a mix.
3. Revisions. Ask how many rounds are included and what counts as a revision. Two or three rounds of notes is a reasonable expectation for a mix.
4. Stems and deliverables. Will you get instrumental versions, a clean version, and stems (grouped audio of drums, bass, vocals and so on) for sync licensing, remixes or live shows? Stems are far cheaper to export now than to recreate later.
5. Communication. The best engineers ask for reference tracks and talk through your vision before they touch a fader.

Mixing typically costs more than mastering because it takes much longer. That's normal. Be wary of any package where mixing is suspiciously cheap and mastering is the headline feature. It usually means the hard part is being rushed.

How to prepare your session so you don't waste money

Half of a great mix is what you hand over. Engineers bill for time, and time spent untangling a messy session is time not spent making your song better. Before you send anything off:

1. Export every track from the same start point. Each file should start at bar one, even if the instrument doesn't come in until the second chorus. That way everything lines up perfectly.
2. Name your files clearly. "Lead Vox," "Kick In," "Bass DI." Not "Audio 14 final final."
3. Leave headroom. Make sure nothing is clipping in the red. Quieter clean files are always better than loud distorted ones.
4. Tell them what's intentional. If that crunchy distorted vocal is the sound, say so, or it might get cleaned up.
5. Send two or three reference tracks. Songs in your genre that sound the way you want yours to feel. This is the fastest way to get on the same page.
6. Include your rough mix. Even a messy demo balance shows the engineer what you've been hearing in your head.

For the master, send a mix with no limiter on the master bus (or send both versions) and some breathing room in level. The mastering engineer needs headroom to do their job.

A worked example: one single, two paths

Say you're an independent artist in Melbourne with a single ready to go and a fixed budget for post-production.

Path one: you mix it yourself over a weekend, then spend the whole budget on a well-known mastering engineer. The master comes back louder and shinier. But your vocal still sits behind the guitars, and the bass swallows the kick on small speakers. The mastering engineer did what they could, and it's still not right. You release it anyway, and it sounds a notch below everything around it on a playlist.

Path two: you split the budget, most of it to a proper mix engineer and the rest to mastering. You send clean, labelled stems, two reference tracks and your rough mix. After two revision rounds the vocal is clear, the low end is tight, and the chorus lifts. Mastering then has a solid song to polish, and the job is quick and subtle. Same budget, completely different result.

The lesson: the money isn't the variable. The order you spend it in is.

How to judge whether your mix and master are done

You don't need golden ears to check the work. You need a routine. Listen to the final master on:

• Cheap earbuds. Can you still hear the bass line and the vocal clearly?
• Your phone speaker. Does the song still make sense with almost no low end?
• The car. The classic test, because everyone listens there.
• Good headphones. Are there clicks, harsh sibilance or odd noises you missed?
• Next to your reference tracks. Match the volume by ear and switch back and forth. Does yours hold up?

If it translates across all five, you're in good shape. If something jumps out, note the timestamp and the system, and send precise feedback. "The vocal at 1:12 disappears on the phone" is useful. "It needs more vibe" is not.

This isn't only for musicians either. Podcasts, brand anthems, ad soundtracks and video scores all go through the same two stages. If you run a show, why your podcast sounds amateur covers the spoken-word version of these fixes. And if you're building an audio identity for a business, sonic branding only works when the sound itself is finished to this standard.

Getting it done properly

Your song is only heard once for the first time. That first listen, on a playlist, in a feed, in someone's car, decides whether they come back. A rough mix and a rushed master make that moment harder than it has to be.

We handle both stages with the same care, so your release sounds professional on every system it touches. Our music and sound team mixes and masters for artists, brands and creators who want their work to stand up next to anything else out there. If you've got a track that's almost there, tell us about it.

Frequently asked questions

What is the difference between mixing and mastering in simple terms?

Mixing blends all the separate recorded parts of a song, like vocals, drums and bass, into one balanced stereo track. Mastering takes that single finished track and polishes it for release, setting the final tone and level so it sounds consistent on every playback system and alongside other songs. Mixing builds the song. Mastering prepares it to compete.

Can I skip mastering if my mix sounds good?

You can, but it's a risk. Mastering adds a fresh, objective listen in an accurate room, catches problems you've stopped hearing, sets appropriate loudness for streaming and handles delivery formats and metadata. A mix that sounds great in your room can still sound thin or harsh elsewhere. For any serious release, mastering is worth the relatively small extra cost.

Should the same engineer mix and master my song?

It can work, especially when budget is tight and the engineer is experienced at both. The traditional argument for separating them is fresh ears: a mastering engineer who hasn't lived inside the mix hears it the way a listener will. If one person does both, ask them to take a break between stages so they come back with a clear head.

How long does mixing and mastering take?

A single song's mix usually takes anywhere from a day to several days of work, plus time for your revision rounds. Mastering is faster, often a day or two per song, including any small revisions. An EP or album takes longer because every track needs to sit together. Build this time into your release plan before you lock in a date.

Usually, they are not. Omnipresence is not a budget. It is a strategy, and a small, sharp brand can manufacture it deliberately.

Here's how, in one sentence: pick a narrow audience, make one strong piece of content at a time, cut it into every format your audience uses, and keep showing up in front of those same people on a steady rhythm through organic posts, email and cheap retargeting. Do that and a small budget buys the feeling of a big brand.

Presence beats reach

Most marketers obsess over reaching as many new people as possible, once. That is backwards. Being seen once by a hundred thousand strangers does almost nothing. Being seen seven times by the right thousand people makes you feel like the obvious choice. Familiarity is trust, and trust is what closes.

Think about your own buying behaviour. You rarely buy from a brand the first time you see it. You notice it, forget it, see it again in a different place, hear someone mention it, see it again. Somewhere along the way it stops being a stranger. When the moment comes to buy, it's the name that feels safe.

That's the whole mechanic. Not persuasion in one hit. Accumulated familiarity across many small touches. And the maths of that favours small brands, because a thousand people is cheap to reach repeatedly, while a hundred thousand is expensive to reach even once.

You do not need to be everywhere. You need to be everywhere your best customer already looks.

Define the corner you're going to own

Omnipresence starts with a narrow definition of "everyone." If your audience is "small businesses in Australia," you can't afford to be everywhere they look. If it's "physiotherapy clinic owners in Brisbane with one to three locations," suddenly you can.

Write down, as specifically as you can:

• Who they are. Role, industry, location, size, life stage. Whatever actually defines your best customers.
• Where they spend attention. Which platforms, which newsletters, which podcasts, which groups, which events. Ask your current customers directly. The answers are often surprising.
• What they worry about. The problems that keep them up at night. Those become your content topics.
• What makes them buy. The trigger moments. A new hire, a bad month, a season, a regulation change.

That list is your map. Everything that follows is about showing up on that map, over and over, for the same people.

The omnipresence playbook

• Pick a tight audience, narrow enough that showing up repeatedly is affordable.
• Create once, repurpose ten times, one idea becomes a post, a reel, an email, a short, a story.
• Retarget relentlessly, the cheapest attention you can buy is from people who already met you.
• Show up on a rhythm, not in bursts, consistency is what builds the feeling of inevitability.
• Concentrate, do not scatter, own two channels completely before touching a third.

Each of those deserves unpacking, because the details are where small brands win or lose.

Create once, publish everywhere

The single biggest reason small brands fail at omnipresence is that they try to create fresh content for every channel. That's a treadmill that ends in burnout within a month.

Flip it. Start with one substantial "pillar" piece each week or fortnight. A short video, a podcast episode, a detailed article, a talk. Then cut it down:

1. Record or write the pillar. One strong idea, explained properly, aimed squarely at your audience's worries.
2. Pull three to five short clips for vertical video on the platforms your audience uses.
3. Turn the key points into a carousel or a text post for the more professional platforms.
4. Write an email to your list with the best insight and a link to the full piece.
5. Pull quotes and stills for stories and graphics.
6. Use the best-performing clip as an ad to your retargeting audience.

One idea, ten appearances, one production effort. We call this an engine rather than a calendar, and we explain the difference in why you should build a content engine, not a content calendar. If video is your pillar, the multiplication gets even stronger: one shoot can feed ninety days of content when it's planned properly.

Why this works when budgets do not

A giant spends to blanket an entire country. You do not need the country, you need your corner of it. By pointing modest resources at a narrow, well-chosen audience and hitting them repeatedly across the places they already spend time, you buy the perception of a giant at a fraction of the cost.

The secret weapon is retargeting. Paid ads to cold audiences are expensive because you're paying to interrupt strangers. Paid ads to people who've already watched your video, visited your site or opened your email are far cheaper, because the audience is small and the platform doesn't have to find them. A modest daily spend can keep you in front of your warm audience almost every day.

There's a right and a wrong way to do that, though. Show the same ad to the same person twenty times and familiarity turns into irritation. The trick is sequencing and frequency caps, which we cover in detail in retargeting funnels that persuade without stalking.

A worked example: the local studio that feels massive

Here's a hypothetical. Say you run a boutique pilates studio on the Central Coast with a marketing budget of about $1,500 a month. Your audience is women aged roughly 30 to 55 within a twenty-minute drive, particularly new mums and people coming back from injury.

Here's what an omnipresence plan could look like:

• Weekly pillar: a short, well-lit video of the founder explaining one common problem, for example why lower back pain flares after sitting all day, with a simple exercise.
• Organic: that video cut into three vertical clips, a carousel and a set of stories across the two platforms your members say they actually use.
• Email: a short weekly note to the list of past trial members and enquiries, with the tip and a link to book.
• Retargeting: most of the ad budget spent showing the best clips, member stories and the intro offer only to people who've watched your videos, visited the site or engaged with the page. Frequency capped, creative rotated.
• Local touchpoints: a partnership with a nearby physio and a café noticeboard, both pointing to the same intro offer.

To someone in that twenty-minute radius, the studio now shows up in their feed, their inbox, their physio's waiting room and their morning coffee stop, all within a couple of weeks. They'll assume it's the busiest studio in the area. The budget didn't grow. The focus did.

What it costs and how to measure it

Omnipresence costs more in consistency than in cash. The real budget is time: a few hours a week to produce the pillar, a system to repurpose it, and the discipline to keep going when early numbers look small.

Rhythm matters more than volume. Posting twice a week, every week, for six months beats posting daily for three weeks and then disappearing. Pick a cadence you can sustain on your worst week, not your best one, and protect it. Batch the work: one production session, one repurposing session, one scheduling session. When the system runs on a fixed day each week, it survives busy periods, holidays and sick days, and the audience never gets the gap that lets them forget you.

For measurement, stop staring at reach and follower counts. Those grow slowly by design. Watch these instead:

• Warm audience size. How many people have watched, visited or engaged in the last 30 to 90 days. This is the pool you're building.
• Branded search. More people searching your name means your presence is sticking.
• "I see you everywhere." Track how often new enquiries say it. It's the best qualitative sign the strategy is working.
• Email list growth and open rates. Email is the channel you own, which is why email deserves more respect in any omnipresence plan.
• Enquiries and sales from the warm audience. The final test. Are the people you keep showing up for actually buying?

The mistakes that make small brands look small

• Spreading across six platforms at once. Thin, inconsistent presence everywhere feels smaller than strong presence in two places.
• Posting in bursts. Three posts a day for a week, then silence for a month. The audience forgets you in the gaps.
• Only running cold ads. Paying to meet new people and never following up is how budgets disappear.
• Changing the message every week. Repetition is how familiarity forms. Say the same core thing in many ways.
• Inconsistent visual identity. If your posts, ads and emails all look different, people don't connect them to one brand, and the familiarity never builds.

What to do this week

1. Write a one-paragraph description of your tightest profitable audience.
2. Ask five recent customers where they spend their attention online.
3. Pick the two channels that come up most and commit to them for ninety days.
4. Make one pillar piece and cut it into at least five assets.
5. Set up a retargeting audience of recent video viewers and site visitors, with a small daily budget and a frequency cap.

Out-strategise, don't outspend

We engineer omnipresence for ambitious brands that refuse to look small, one content engine, feeding every channel, aimed with precision at the people who matter. If your competitors feel bigger than you when they are not, it is because they out-strategized you, not outspent you. Let's flip that.

Our digital marketing team builds the full system: audience definition, the pillar content, the repurposing, the email and the retargeting, all running on one rhythm so your brand keeps turning up in the right places.

Frequently asked questions

How much budget do I need to be omnipresent?

Less than most people think, because omnipresence is about narrow focus rather than broad reach. The key costs are consistent content production and a modest retargeting budget aimed at people who already know you. A tightly defined local or niche audience can be reached repeatedly on a small monthly spend. The bigger investment is usually time and consistency, not advertising dollars.

Which social media platforms should a small business focus on?

The ones your best customers already use, not the ones that are trending. Ask recent customers where they spend their time online and pick the two platforms that come up most. Commit to those properly for at least ninety days before adding a third. Strong, consistent presence on two channels beats a thin, irregular presence on six.

How long does it take for an omnipresence strategy to work?

Expect early signs within a few weeks, such as a growing warm audience and people saying they keep seeing you. Meaningful results in enquiries and sales usually take a few months of consistent effort, because familiarity builds through repeated exposure over time. Brands that quit after a month rarely see the payoff, since the compounding only starts once people have encountered you several times.

What is the difference between omnipresence and spamming people?

Omnipresence means showing up with useful, varied content across the places your audience already spends time. Spam is the same message repeated too often in the same place. Rotate creative, cap ad frequency, give value in most touchpoints and exclude people who have already bought. Done well, people feel like they keep running into a helpful brand, not like they are being followed.

So what makes product photography actually sell? Three things. Light that proves quality, a set of shots that answers every question a buyer would ask if the thing were in their hands, and a consistent visual system that makes your whole range look like one premium brand instead of a garage sale.

Get those right and your photos do the job a great salesperson does in a shop. Get them wrong and it doesn't matter how good the product is, because nobody online will ever find out.

Your photo is standing in for the shop floor

Think about how people buy in person. They pick the thing up. They turn it over. They check the stitching, feel the weight, look at it next to something they already own. They squint at the label. Every one of those moments removes a doubt.

Online, none of that happens. The buyer has a thumb, a phone screen and a few seconds of patience. Every doubt you would have removed in person is now sitting there unanswered, and an unanswered doubt is one of the ways a cart gets abandoned.

That's the real job of product photography. Not to look nice. To do the physical inspection on the buyer's behalf, so that by the time they tap "Buy" they feel like they've already held it.

Most product photos fail that test. They show the product once, from the front, on a white background, in flat light, and leave the buyer to guess the rest. Guessing is friction. Friction kills sales.

Light tells the truth about quality

The difference between "cheap" and "premium" is often just lighting. Soft, controlled light reveals texture and form, harsh, flat light flattens everything into something forgettable. Great product photography sculpts with light to make materials look as good as they feel in hand.

Here's what that means in practice:

• Soft light for skin, fabric and food. A large diffused source wraps around the product and gives gentle gradients. It makes cotton look soft and bread look fresh.
• Hard, directional light for texture. Leather grain, brushed metal, raw timber and woven materials come alive when light skims across them from a low angle. The shadows are where the texture lives.
• Controlled reflections for glass, metal and gloss. Shiny products don't show light, they show whatever is around them. Pros shape the reflections with large panels so a bottle gets a clean, elegant highlight instead of a mirror image of the studio.
• Accurate colour, every time. If the navy jumper arrives looking black, that's a return, a bad review and a customer who won't trust your photos again. Colour accuracy isn't a creative choice. It's customer service.

The phone-on-the-kitchen-bench shot fails here almost every time. Not because phones are bad cameras, but because overhead downlights and window glare were never designed to flatter a product. The light decides whether your thing looks like a treasure or a trinket.

People don't buy products. They buy the better version of themselves the product promises, and the photo has to sell that.

Show it in their world

One hero image is a start. It is not a sales page. The listings that convert tell a short visual story, and each frame has a specific job.

• Hero shots establish the product as the star, clean, detailed, aspirational.
• Context shots let the buyer imagine it in their life, their hand, their space.
• Detail shots answer the silent questions, the stitching, the finish, the proof of craft.

Add two more to that list and you cover almost everything a buyer wants to know:

• Scale shots. The product next to a hand, on a shelf, on a person, beside an everyday object. "How big is it actually" sits behind plenty of returns, and a single honest scale shot answers it.
• What's in the box. Everything the buyer receives, laid out. It sets expectations and makes the purchase feel like good value before it arrives.

Context is where most brands leave money on the table. A lamp on white is a lamp. The same lamp glowing on a bedside table next to an open book on a rainy evening is a feeling, and the buyer is purchasing the feeling. That's the "better version of themselves" from the quote above, made visible.

If you want the flagship image to carry real weight, it's worth reading our breakdown of what separates a hero shot from a snapshot. The same principles apply whether the frame is still or moving.

The shot list that answers every buyer question

Before any shoot, write down the questions your customers actually ask. Pull them from your inbox, your reviews, your returns reasons and your sales team. Then build a shot for each one. Here's the process we'd follow for a single product:

1. List the objections. Is it well made? Will it fit? Is the colour true? What does it look like on? Is it worth the price?
2. Map one image to each objection. Craftsmanship gets a macro detail. Fit gets a scale shot. Price gets a premium hero and a styled lifestyle frame that makes the value obvious.
3. Order the gallery like a sales conversation. Hero first to stop the scroll, then context to create desire, then details to remove doubt, then scale and contents to close.
4. Plan the crops before you shoot. Your images will live in square marketplace tiles, vertical stories, wide website banners and tiny search thumbnails. Shoot with room around the product so one frame can serve every placement.
5. Shoot the whole range in one session. Same lighting, same angles, same styling rules. That's how you get consistency without paying for it twice.

Notice what's missing from that list: "make it look cool." Cool is a by-product. The goal is a buyer with no unanswered questions.

Consistency builds the brand

A cohesive visual language across every product turns a catalog into a brand. We build photography systems, consistent lighting, angles, and styling, so your storefront looks intentional, premium, and unmistakably yours.

Scroll a store where every product was shot on a different day, in a different light, by a different person. It feels like a market stall. Now scroll one where every image shares the same background tone, the same shadow direction, the same camera height. It feels like a brand you can trust with your card details.

A photography system is a simple document that makes this repeatable:

• Lighting setup, noted precisely enough that next season's products match this season's.
• Standard angles, for example front, three-quarter, back and detail, always in the same order.
• Background and surface palette, chosen to suit your brand colours rather than whatever was in the studio.
• Props and styling rules, what's allowed in frame, what isn't, and how much space the product gets.
• Retouching standard, clean up dust and blemishes, never change what the product actually is.

That last rule matters more every month. Generative tools can now restyle or invent product shots in seconds, and it's tempting to skip the shoot. We've written about why fake product photos can torch your brand: buyers spot the tells, and the moment the real item doesn't match the image, you've lost more than a sale. Real photography of your real product has quietly become a trust signal.

A worked example: one product, one afternoon

Here's a hypothetical to make it concrete. Say you run a small Australian leather goods brand selling a handmade wallet online. Your current listing has one photo: the wallet, closed, on a white sheet, shot on a phone under the kitchen light. It looks fine. It sells slowly, and a few customers have asked whether it fits cards and cash comfortably.

A proper shot list for that wallet might look like this:

1. A hero on a warm-toned surface, lit from the side so the grain and the hand stitching throw tiny shadows.
2. The wallet open and loaded with cards and a folded note, answering the capacity question before it's asked.
3. A macro of the stitching and the burnished edge, the proof of handmade craft.
4. The wallet sliding into a back pocket, for scale and for the feeling of carrying it.
5. A lifestyle frame on a café table next to keys and a flat white, the version of the buyer's day they're picturing.
6. The wallet in its packaging, so the gift buyer knows it arrives looking like a gift.

Six images, each answering a different doubt. Same light across all of them, so they belong together. Nothing about the product changed. What changed is that a stranger scrolling on a phone can now inspect it as thoroughly as someone standing at your market stall. That's the whole game.

What it takes, and where the money goes

Good product photography isn't cheap, and it isn't meant to be the cheapest line in your marketing budget. What you're paying for:

• Lighting control. The single biggest difference between amateur and professional product work, and the hardest to fake.
• Styling and art direction. Choosing surfaces, props and composition that fit your brand and your buyer.
• Retouching. Removing dust, fingerprints and seams in the backdrop without changing the product.
• A system you can reuse. The real value. Once the look is defined, new products slot in without reinventing anything.

The smarter way to think about cost is per image, per year. One well-planned shoot can supply your store, your marketplace listings, your ads, your emails and your social feed for months. Pair it with short product clips from the same session and the value multiplies again. The photos also double as your ad creative, and the scroll-stopping rules we cover in the psychology of the scroll-stopping thumbnail apply directly to your product hero.

How to tell if your photos are working

Don't judge product photography by whether you like it. Judge it by what buyers do. Watch these:

• Product page conversion rate. The cleanest signal. Compare before and after a reshoot on the same product with similar traffic.
• Return reasons. "Not as pictured," "smaller than expected" and "different colour" are photography problems wearing a logistics costume.
• Pre-sale questions. If your inbox keeps asking the same thing, there's a missing shot.
• Ad click-through rate. When the product image is the ad, a stronger hero shows up here first.
• Gallery engagement. If your store platform shows how many images people swipe through, a buyer who looks at five is usually closer to buying than one who looks at one.

Common mistakes that quietly cost you sales

• One photo per product. You've given the buyer a question instead of an answer.
• Over-editing. Plastic-smooth surfaces and punched-up colour look impressive on screen and disappointing on the doorstep.
• Clutter. Props should support the product, never compete with it.
• Inconsistent backgrounds across the range. It makes the whole store feel unplanned.
• Shooting only for the website. Then scrambling to crop for socials and ads, and losing the product off the edge of the frame.
• Ignoring mobile. Most shoppers will see your image at thumbnail size first. If the product isn't readable small, the big version never gets opened.

Make the photo do the selling

Here's the uncomfortable truth. A lot of brands blame their ads, their pricing or their platform when the real problem is sitting right there in the gallery. The product is great. The photos just never let anyone find that out.

Fix the light, answer every question with a frame, and lock it all into a system so your range looks like one confident brand. That's the work our photo and video team does: shot lists built around buyer objections, lighting that proves quality, and a visual system you can keep using long after the shoot wraps. If your product deserves better than a kitchen-bench snapshot, let's give it the photos that make people tap "Buy."

Frequently asked questions

How many product photos should each listing have?

Enough to answer every question a buyer would ask in person. For most products that means a hero shot, one or two context or lifestyle frames, a close-up detail, a scale shot and a what's-in-the-box image. Five to seven is a sensible range. One photo almost always leaves doubts unanswered, and unanswered doubts lead to abandoned carts and returns.

Is a white background still the best choice for product photos?

White is ideal for your main hero and for marketplaces that require it, because it's clean and consistent. It shouldn't be your only look. Context and lifestyle shots on surfaces that suit your brand are what create desire. The winning approach is usually a white or neutral hero for clarity, then styled images further along the gallery to sell the feeling.

Can I shoot product photos on my phone?

You can get usable results on a modern phone if you control the light, use a stable surface or tripod, and keep backgrounds consistent. The camera is rarely the problem. Lighting, styling and colour accuracy are. For a small range, phone photos in good diffused light beat studio photos taken carelessly. For a premium brand, professional lighting control is usually where the difference shows.

Should I use AI to generate or edit product images?

Use AI for tasks that don't change the truth of the product, like removing dust, extending a background for a new crop or concepting before a shoot. Avoid generating the product itself or altering its colour, shape or details. If the item that arrives doesn't match the image, you risk returns, bad reviews and a customer who stops trusting every photo you publish.

Google works fine. The problem is almost always the same single thing: intent. Fix that, and the exact same budget stops bleeding and starts buying customers.

Put simply, Google Ads wastes money when you pay for searches from people who aren't ready to buy, then send them to a page that doesn't match what they searched. Target high-intent searches, block the wrong ones, and match every ad to a page built for that exact search, and the same platform starts producing customers instead of invoices.

Not all searches are created equal

Someone typing "what is cold plunge therapy" is curious. Someone typing "cold plunge tub buy near me" is reaching for a wallet. These are wildly different humans, and paying the same to reach both is how you set money on fire. Most accounts pour budget into curiosity searches that will never convert, then blame the platform for the empty pipeline.

You are not buying clicks on Google. You are buying intent. Buy the wrong intent and every click is a small donation.

It helps to sort searches into three buckets:

• Learning searches. "What is," "how does," "ideas for," "why is my." These people want information. They might buy one day, but not today. They're better served by content and SEO than by paid clicks.
• Comparing searches. "Best," "vs," "reviews," "cost of." These people are weighing options. Worth bidding on selectively, with a page built to win the comparison.
• Buying searches. "Near me," "buy," "book," "quote," "hire," "emergency," a specific product name. These people have a problem and money to fix it. This is where most of your budget belongs.

The fewer people search in general, the more this matters. The ones who still type into Google are often further along. We dig into that shift in fewer people are searching, the ones left are ready to buy.

Where the money leaks

• Bidding on broad, curious searches instead of ready-to-buy ones.
• No negative keywords, so you pay for searches that will never buy.
• Sending every click to the homepage instead of a page that matches the search.
• Optimising for clicks instead of customers, cheap traffic that never converts.
• Ignoring the message match between what they searched, what your ad promised, and what the page delivers.

Two more leaks that don't show up until you look:

• Loose match types left running unchecked. Broad match can find good searches you'd never have thought of. It can also match you to searches that are only loosely related. Without a weekly look at the search terms report, you won't know which is happening.
• Counting the wrong conversions. If a page view, a click on your phone number or a newsletter sign-up counts as a "conversion," Google will happily find you more of those. Your dashboard looks healthy while your sales team starves.

How to find the leak in your own account

You don't need to be a Google Ads specialist to spot where your money is going. Block out one hour and do this:

1. Open the search terms report for the last 30 days. This shows what people actually typed, not the keywords you chose. Sort by cost.
2. Mark every search as buy, compare, learn or irrelevant. Be honest. "Free," "jobs," "DIY," "course," "salary" and "how to" are usually red flags for a business selling a service.
3. Add up the spend on learn and irrelevant searches. That's your leak, in dollars.
4. Check your conversion settings. Under goals, look at exactly what counts as a conversion. If it's anything short of a real enquiry, call or sale, fix that first.
5. Check where each ad sends people. Click your own ads (from a preview, not a live search) and ask: does this page answer exactly what the searcher asked for, above the fold?

If you want a broader version of this audit across every channel, not just Google, we lay one out in the audit that finds your wasted ad spend.

The fix is intent, end to end

Winning on Google means chasing buyers, not browsers: targeting high-intent searches, ruthlessly filtering out the wrong ones, and matching each ad to a page that delivers exactly what the searcher wanted. Do that and your cost per customer can fall through the floor, same platform, same budget, completely different result.

Here's what "end to end" means, link by link:

• The keyword. Built around buying language and specific services, grouped tightly so each group covers one intent.
• The negatives. A living list of words and phrases you never want to pay for, added to every week from the search terms report.
• The ad. Repeats the searcher's words back to them, states the offer and gives one reason to choose you over the other three ads on the page.
• The landing page. Continues the exact promise of the ad. Same words in the headline, proof close by, one clear action. Our breakdown of a landing page that prints money covers what goes on it.
• The conversion. A real enquiry, booked call or purchase, ideally with sales outcomes fed back into Google so it learns which clicks became customers.

A worked example: the emergency plumber

Say you run a plumbing business in Adelaide spending $4,000 a month on Google Ads. You're bidding on "plumber," "plumbing," and "blocked drain" on broad match. Every ad goes to your homepage. Conversions are counted as "any visit to the contact page." Google reports plenty of conversions. Your phone doesn't ring much.

You open the search terms report and find money going to "plumber salary," "plumbing apprenticeship," "how to unblock a drain yourself" and "plumbing supplies near me." None of those people want to hire you.

So you rebuild. You split campaigns by intent: emergency searches ("emergency plumber Adelaide," "burst pipe plumber now") and planned work searches ("hot water system replacement quote"). You add negatives for salary, jobs, apprenticeship, DIY, how to and supplies. You build two landing pages, one for emergencies with a big tap-to-call button and your response time front and centre, and one for hot water with pricing guidance and a quote form. You change the conversion to real calls over a set length and submitted quote forms.

Same $4,000. Fewer clicks, most likely, because you've stopped paying for the junk. But every click now comes from someone with a problem you fix, landing on a page that fixes it. That's how the maths flips. (The business and figures here are hypothetical, but these are the exact levers.)

How to measure whether it's working

Most Google Ads reports lead with clicks, impressions and click-through rate. Those numbers tell you whether people noticed your ad. They don't tell you whether you made money. Once intent is fixed, judge the account on three numbers instead.

• Cost per qualified lead. Total spend divided by the enquiries your team would actually want to take. Not every form fill. The good ones.
• Cost per customer. Total spend divided by the people who paid you. This is the number that matters most, and it needs your sales records, not just the Google dashboard.
• Customer value versus cost. What a customer is worth to you over the first year compared with what it cost to win them. If a customer is worth $2,000 and costs $300 to acquire, you have a machine. If they're worth $400 and cost $500, you have a hobby.

To get there, you need a simple way to connect ad clicks to sales. That can be as basic as asking every new enquiry "how did you find us?" and logging it in a spreadsheet, or as thorough as importing closed sales from your CRM back into Google. Either is better than trusting the default dashboard. And once Google learns which clicks turned into paying customers, its bidding gets much smarter about finding more of them. (The dollar figures above are illustrations, not benchmarks. Your numbers will depend entirely on your margins.)

What about Google's AI campaigns?

Google keeps pushing advertisers toward automated campaign types that decide more of the targeting for you. Automation isn't the enemy. Automation pointed at the wrong goal is. If your conversions are junk and your landing pages are generic, the machine will get very efficient at buying you junk.

Give it good inputs, like real conversion data, clear negatives, strong landing pages and ad copy grounded in buyer intent, and it has something worth optimising toward. We explain what's changing, and how to protect yourself, in Google is moving your campaigns to AI Max.

Common mistakes after the fix

• Cutting budget the moment clicks drop. Fewer, better clicks is the goal. Judge on cost per customer, not cost per click.
• Setting negatives once and forgetting them. New junk searches turn up every month. The search terms report is a weekly habit.
• Building one landing page for everything. One page per intent is the minimum. One per service is better.
• Not answering the phone. It sounds obvious. But high-intent searchers call the next business if you don't pick up. Your ad spend is only as good as your response time.
• Judging too early. Give the rebuilt account a few weeks of clean data before you decide it's working or not.

You don't have a Google problem

We rebuild Google campaigns around intent, so every dollar chases someone with their wallet already half out. That's how our advertising team approaches every account: find the leak, fix the conversion data, rebuild around buying searches and match every ad to a page that closes.

If Google feels like a money pit, you do not have a Google problem. You have an intent problem. Let's fix it.

Frequently asked questions

Why am I getting clicks but no sales from Google Ads?

Clicks without sales usually mean you're paying for the wrong searches, sending people to a page that doesn't match what they searched, or counting the wrong actions as conversions. Check the search terms report for irrelevant or research-only searches, add negative keywords, point each ad at a matching landing page and make sure Google only counts real enquiries or sales.

What are negative keywords in Google Ads?

Negative keywords are words or phrases you tell Google never to show your ads for. If you sell a plumbing service, adding "salary," "jobs," "DIY" and "course" as negatives stops you paying for people looking for work or free instructions. Build the list from your search terms report and add to it every week, because new irrelevant searches keep appearing.

Should I send Google Ads traffic to my homepage?

Usually not. Your homepage speaks to everyone, so it rarely answers the specific search that triggered the ad. A dedicated landing page that repeats the searcher's words, shows relevant proof and offers one clear next step will almost always convert better. Build at least one page per intent, such as emergency work versus planned quotes.

How much should a small business spend on Google Ads?

Start with what you can afford to test for a couple of months without panic, then let cost per customer guide you. A smaller budget focused on high-intent searches often beats a big budget spread across everything. Once you know what a customer costs and what one is worth, scaling becomes a maths decision rather than a guess.

If you're losing deals to competitors who aren't obviously better than you, the most likely cause is your message. Prospects can't quickly tell what you do, who it's for and why it's the right choice, so they pick the business they understood first. Fix the message and you often fix the pipeline, without changing the product or the budget.

Confusion is the silent deal-killer

A prospect lands on your site or hears your pitch and, within a few seconds, their brain asks one question: "What is this, and is it for me?" If the answer is not immediate and obvious, they do not email to clarify. They leave. Confused people do not buy, they bounce, and you never even know they were there.

That's what makes a messaging problem so dangerous. It doesn't show up as a rejection. It shows up as silence. Fewer enquiries than you expected. Quotes that go nowhere. Prospects who "need to think about it" and never come back. It feels like a market problem or a pricing problem, so you cut prices, spend more on ads or redesign the logo. None of that helps if people still can't work out what you do.

You do not lose sales because people decided no. You lose them because people could not quickly decide yes.

How to tell if your message is the problem

Most business owners are too close to their own message to see it clearly. You know exactly what you mean, so your words make sense to you. Here are the warning signs that they aren't landing for anyone else:

• Prospects ask what you do after reading your website. If your homepage needs a follow-up explanation, it isn't doing its job.
• You attract the wrong enquiries. Tyre-kickers, people wanting services you don't offer, budgets that don't fit. Unclear messages attract unclear leads.
• Your team describes the business differently. Ask three staff members what you do in one sentence. If you get three different answers, so do your customers.
• You win when you get in the room, and lose everywhere else. That means the value is real but the words aren't carrying it without you there to translate.
• Your copy could belong to any competitor. Swap your logo for theirs. If nothing else needs to change, you have no message, just category noise.

Two or more of those and it's time to stop blaming the competition.

Clarity is a competitive weapon

When your message is instantly clear, who you help, what you do, why it matters, why you, something powerful happens. The right people feel it immediately. They think "this is exactly what I have been looking for." You did not need to be better than the competition. You just needed to be the one they actually understood.

• Lead with the outcome your customer wants, not the mechanics of what you do.
• Cut the jargon. If a smart outsider cannot follow it, it is costing you sales.
• Say who it is for, and who it is not. Specificity signals fit.
• Make the next step obvious. Never make people hunt for how to buy.
• Repeat the core message everywhere until it is unmistakable.

Each of those principles attacks a different kind of friction. Leading with the outcome answers "why should I care?" Cutting jargon answers "what does that mean?" Saying who it's for answers "is this for me?" A clear next step answers "what do I do now?" And repetition makes sure the answer is the same everywhere a prospect looks. This also connects directly to positioning. If you haven't decided exactly who you serve and why you're different, no amount of wordsmithing will fix it. We cover that foundation in why positioning comes before the logo.

The five-second message framework

You need a core message a stranger can understand in five seconds. Here's a simple way to build one:

1. Name the customer. Be specific. "Small businesses" is not a customer. "Owner-run trade businesses with five to twenty staff" is.
2. Name their problem in their words. Use the language customers actually use on calls and in emails, not the technical name you use internally.
3. Name the outcome. What does life look like after you've helped them? More time, more profit, less risk, fewer headaches.
4. Give one reason to believe. Your method, your specialisation, your guarantee, your track record. Just one, the strongest.
5. Give the next step. Book a call, get a quote, see pricing. One action, clearly labelled.

Put it together and you get something like: "We help owner-run trade businesses stop chasing unpaid invoices, so you get paid on time without awkward phone calls. Book a free cash-flow check." Not poetry. Just clear. And clear wins.

A quick test: read your headline to someone outside your industry, then ask them to repeat back what you do and who it's for. If they can't, rewrite it. If they can, you're ready to use it everywhere.

Where your message leaks sales

A clear core message isn't much use if it only lives in a strategy document. It has to show up at every point where a buyer is making up their mind:

• Your homepage headline. The single most important sentence in your business. It should state the outcome and who it's for, not a vague slogan.
• Service and landing pages. Each one should answer the questions a buyer has at that moment, in the order they have them. Our breakdown of the anatomy of a landing page that prints money walks through the structure.
• Ads. An ad with a muddy message pays for clicks from confused people.
• Sales calls and proposals. If every salesperson pitches differently, your prospects hear a different business every time.
• Your social bios, email signatures and Google Business Profile. Small places, big impact, because they're often the first thing people see.

Your website carries most of the load here, because it's where prospects go to check you out when you're not in the room. It should be selling for you around the clock, and we've written about why your website is your best salesperson when it's doing that properly.

The fix is cheaper than you think

Here is the good news buried in all this: a messaging problem is one of the cheapest, fastest things to fix in a business. You do not need a new product or a bigger budget. You need the right words in the right order, and the results can be immediate.

Compare that with the alternatives. Developing a new product takes months. Increasing ad spend costs money every single day. Hiring more salespeople is expensive and slow. Rewriting your message is a focused piece of thinking and writing that then improves the return on everything else you already spend. Every ad, every visit, every sales call works harder when the words are right.

Start this week. Pull up your last ten enquiries, sales call notes or customer emails and highlight the exact phrases people used to describe their problem. Those words, not your internal jargon, are your raw material. Then rewrite only your homepage headline and the line beneath it using the five-second framework above. Measure enquiry volume and enquiry quality for a month before and after. It's the smallest possible change with the biggest possible reach.

A worked example: the accountant nobody understood

Here's a hypothetical. Say you run a small accounting firm in Perth. Your homepage headline reads: "Integrated financial solutions for a changing world." Your services page lists tax, BAS, bookkeeping, advisory and "strategic financial planning." You're good at what you do and your existing clients love you. But new enquiries are slow, and most of them want the cheapest possible tax return.

You sit down and work through the framework. Your best clients turn out to be hospitality owners with one to three venues. Their real problem isn't tax. It's not knowing whether they're making money week to week until it's too late. The outcome they want is confidence in their numbers.

So you rewrite. New headline: "Accounting for cafe and restaurant owners who want to know they're profitable every week, not once a year." You add a short section that speaks to the exact moments they struggle with: rising supplier costs, rosters that blow out, quiet months. You make the next step a free weekly-numbers review.

Nothing about the firm changed. Same people, same skills, same prices. But now the right prospect reads that headline and thinks, "That's me." That's the whole point. The numbers in this scenario are made up, but the mechanism is real: specificity attracts fit.

Common messaging mistakes

• Trying to appeal to everyone. A message for everybody lands with nobody. Choosing a focus feels scary, but it's what makes the right people choose you.
• Talking about yourself first. "We are a passionate team with twenty years' experience" is about you. Customers care about their problem first.
• Leading with being the best. "Best quality, best service" is what every competitor says too. We explain why claiming to be the best quietly kills your business.
• Being clever instead of clear. A witty tagline nobody understands costs more than a plain one everyone does.
• Changing the message every month. Consistency builds recognition. You'll be bored of your message long before your market has even noticed it.

Become the clearest voice in your market

We turn muddy, forgettable messaging into words that land in five seconds and make the right people lean in. If you suspect you are losing sales to confusion rather than competition, you almost certainly are. Let's make you the clearest voice in your market. Our consulting work starts exactly here, with the words that decide whether a stranger becomes a customer.

Frequently asked questions

How do I know if my marketing message is clear?

Show your homepage to someone outside your industry for five seconds, then ask what you do, who it's for and what they should do next. If they can't answer all three, your message isn't clear yet. Other signs include lots of off-target enquiries, staff describing the business differently and prospects who only get it once you explain in person.

What should my homepage headline say?

It should state the outcome you deliver and who you deliver it for, in plain words your customers would use. Avoid slogans, jargon and claims every competitor makes. A supporting line can add how you do it or a reason to believe, and a clear button should tell visitors the next step, such as booking a call or getting a quote.

Is messaging the same as branding?

They're connected but not the same. Branding covers how your business looks, sounds and feels overall, including visuals, tone and reputation. Messaging is the specific words that explain what you do, who it's for and why it matters. Good messaging sits on top of clear positioning, and it's usually the fastest part of your brand to fix.

How long does it take to fix a messaging problem?

The core thinking can often be done in a focused workshop or a few sessions, and the rewritten homepage, key pages and pitch can follow within weeks. The bigger job is rolling the message out consistently everywhere prospects see you. Results can show up quickly, because every visit and conversation starts benefiting as soon as the new words go live.

So is writing a book worth it if it never sells many copies? Yes, if you treat it as a business asset instead of a retail product. A well-positioned book wins you clients, speaking invitations and pricing power long before it earns you a cent in royalties, because its real job is to make a stranger trust you faster than anything else you can hand them.

This article breaks down exactly how that works, what kind of book actually does it, what it takes to produce one, and how to measure whether yours is paying its way.

Authority you can hand someone

Anyone can claim expertise. A book proves it. It's the difference between saying you know your field and handing a prospect 200 pages that demonstrate it. Instant credibility, in a format they can't ignore.

Think about what a prospect normally has to go on when they meet you. A website that looks like everyone else's. A LinkedIn profile full of the same words your competitors use. A sales call where you both say polite things for thirty minutes. None of that separates you. It all asks them to take your word for it.

A book flips that dynamic. You stop asking for trust and start showing the thinking that earns it. The reader sees how you diagnose problems, what you believe that your competitors don't, and the method you actually use. By the time they reach the last chapter, they've spent hours inside your head. No brochure, ad or pitch deck gets that kind of attention.

A great book turns "who is this?" into "I've been wanting to talk to you."

There's a second, quieter effect. Writing the book forces you to organise what you know. Most experts carry their method around as instinct. Putting it on paper turns instinct into a framework with names, steps and examples, and that framework becomes the backbone of your sales conversations, your workshops and your content for years.

Leverage that scales

• Lead generation: The book attracts exactly the people you want as clients and repels the wrong ones.
• Sales enablement: It pre-sells your philosophy before a single call. Prospects arrive already convinced.
• Speaking and press: "Author of" opens doors that cold outreach never will.
• A platform: Each reader becomes part of an audience you can serve for years.

Here's why each of those compounds. A blog post gets skimmed and forgotten. A sales call reaches one person at a time. A book sits on a desk, gets passed to a business partner, gets quoted in a board meeting. It travels further than you do, and it carries your full argument with it, unedited, every time.

It also changes the conversation you have when a prospect finally reaches you. Instead of explaining who you are, you're discussing chapter four and how it applies to their situation. That's a sales call that starts halfway to yes.

The book that works versus the book that sits in a box

Not every business book does this. Plenty of founders write one, order a few hundred copies, and watch them gather dust in the garage. The difference is almost never the quality of the prose. It's the strategy behind it.

The books that generate business share a few traits:

• They solve one specific problem for one specific reader. "Everything I know about leadership" helps nobody. "How owner-operated trade businesses hire their first manager without losing control" gets read, cover to cover, by exactly the right people.
• They teach the what and the why generously. Hoarding your method doesn't protect you. Readers who understand the method realise how much skill it takes to execute, and they hire the person who wrote it.
• They're short enough to finish. A busy decision-maker will read something they can get through on a flight. They won't start a doorstop.
• They have a clear next step. A reader who finishes and wants more should know exactly where to go, whether that's a diagnostic call, a workshop or a resource page.
• They sound like a person. Stories, opinions and real scars. We've written about why nobody finishes a boring business book, and it's the fastest way to waste the whole exercise.

If you can't describe your ideal reader and the single problem you solve for them in one sentence, don't start writing yet. Start there.

A worked example: the consultant and the 100 copies

Here's a hypothetical to make it concrete. Say you run a small operations consultancy in Brisbane, helping manufacturing businesses cut waste. Your sales cycle is long, your prospects are sceptical, and every proposal ends up in a price comparison with bigger firms.

You write a focused book on the five places manufacturers quietly lose margin, with the diagnostic you use in your first week on site. You don't chase bestseller lists. Instead you do three things with it:

1. You mail a signed copy to a shortlist of operations managers you'd genuinely love to work with, with a short personal note pointing to the chapter most relevant to their business.
2. You give it to every prospect after a first meeting, so the thinking stays on their desk while they compare quotes.
3. You pitch industry events and podcasts on the book's core idea, not on yourself.

Now the maths. Even if only a handful of those people ever reply, one new client on a long engagement could be worth more than the entire cost of producing the book. And every prospect who reads it arrives at the next meeting speaking your language. Those are hypothetical numbers, but the logic is exactly why a small print run can outperform thousands of retail sales.

Notice what the royalties are worth in that scenario. Almost nothing. And it doesn't matter.

What it actually takes to produce one

The biggest objection we hear is time. Founders assume a book means a year of weekends. It doesn't have to. A structured process looks something like this:

1. Positioning first. Nail the reader, the problem and the promise before a word gets written. This single step decides whether the book sells your business or just exists.
2. Outline the argument. Map the chapters as a journey from the reader's current pain to the outcome they want, with your method in the middle.
3. Capture your knowledge. For many experts the fastest route is talking, not typing. Recorded interview sessions pull out the stories and frameworks you'd never think to write down.
4. Draft and edit hard. Turn the raw material into clean, readable chapters, then cut everything that doesn't serve the reader.
5. Design and publish. A professional cover and interior, print and ebook formats, and the distribution decision. Our breakdown of self-publishing versus traditional publishing covers which route fits which goal.
6. Deploy it. The launch plan and the ongoing use of the book in sales, events and content.

If you want the timeline version of this, we've laid out how busy founders actually get a book done in 90 days. The short answer is structure and outside help, not more willpower.

On cost, the honest answer is that it depends on how much you do yourself. Writing it entirely alone costs your time. Working with a ghostwriter or book team costs money but gets it finished. Either way, measure it against the value of the clients it brings in, not against book sales.

It works while you sleep

Write it once and it sells your expertise on a Tuesday at 3am, in a market you've never visited, to a person you'll never meet, until they email asking to work with you.

That's the real leverage. Your calendar is finite. Your book isn't. It can be in a hundred inboxes, on a hundred desks and in a hundred carry-on bags at the same time, each one making your case in full.

It also feeds everything else you publish. Each chapter can become a keynote, a workshop, a newsletter series, a set of short videos. You stop wondering what to post because you've already done the deep thinking. The book becomes the source, and your content becomes the distribution.

How to measure whether your book is paying off

If royalties are the wrong scoreboard, you need a right one. Track these instead:

• Enquiries that mention the book. Add "How did you hear about us?" to your enquiry form and count the answers.
• Sales cycle length. Do prospects who've read it decide faster than those who haven't?
• Close rate and deal size. Authority usually shows up as less discounting and bigger engagements.
• Invitations. Speaking slots, podcast interviews and media requests you didn't have to chase.
• Copies placed with the right people. Ten copies in the hands of ideal clients beat a thousand random sales.

Review these every quarter. A business book is a long game, and the returns tend to arrive as warmer conversations long before they show up as a spike in revenue.

Common mistakes that waste the whole effort

• Writing for everyone. A book with no clear reader gets read by no one.
• Treating the launch as the finish line. The book earns its keep in the months and years after launch day, as a sales and credibility tool. Our piece on why book launches are where most authors die goes deeper on this.
• Skimping on the cover and edit. People absolutely judge a book by its cover, and a cheap-looking one undermines the authority it's meant to build.
• Hiding the method. Vague books built to tease a paid offer feel like a brochure and get treated like one.
• Never mentioning it. If your book isn't in your email signature, proposals, website and bio, it can't do its job.

The business card that does the talking

A business card tells people your name. A book tells them why they should care. It's the only marketing asset people will willingly spend hours with, pay for, keep on a shelf and pass on to someone else.

That's why the smartest founders don't ask whether their book will be a bestseller. They ask whether it will put their thinking in front of the right hundred people. That second question is one you can actually answer, and win.

We help experts turn what's in their head into a book that does the selling for them, from positioning and ghostwriting to design and publishing. If you've been carrying a book around in your head for years, have a look at our books and publishing service and let's get it out where it can work for you.

Frequently asked questions

Is it worth writing a business book if it won't be a bestseller?

Yes. For founders, consultants and coaches, most of the value comes from what the book does for the business: credibility with prospects, speaking and media invitations, shorter sales cycles and better-fit clients. A book that reaches the right few hundred readers can be worth far more than one that sells widely to people who will never hire you.

How long should a business book be?

Long enough to solve one problem properly and short enough for a busy reader to finish. Many effective business books can be read in a few sittings or on a single flight. Length should come from the depth of the method you're teaching, not from a target page count. Cut anything that doesn't move the reader toward the outcome you promised.

Can I write a book if I'm not a good writer?

Absolutely. Your expertise, stories and opinions are the raw material, and those only come from you. Many experts work with a ghostwriter or book team who interviews them, drafts the chapters and edits them into shape. The book still sounds like you and contains your thinking, you just don't have to wrestle every sentence onto the page yourself.

How do I use my book to get clients?

Put it directly in front of the people you want to work with. Send signed copies to ideal prospects, give it after first meetings, mention it in proposals and your email signature, and use its core idea to pitch podcasts and events. Include a clear next step inside the book so readers who want help know exactly how to reach you.

So is royalty-free music actually bad for your brand? Not always, but when it becomes the sound of your brand, yes. A track anyone can license can't make you distinctive, can't build memory, and quietly tells your audience you're the same as everyone else using it. Stock music is fine for filler. It is a terrible foundation.

Most businesses never think about this, because music feels like the last step. The visuals get the budget, the copy gets the debate, and the soundtrack gets whatever was on page one of the library at 4pm on deadline day. Here's why that habit costs you more than you think, and what to do instead.

Familiar in the worst way

Popular stock tracks get used everywhere, by your competitors, by unrelated brands, by that awkward corporate video from 2019. When a viewer half-recognises your music from somewhere else, it quietly undermines you. Instead of feeling unique and premium, your brand feels generic and interchangeable. You cannot build a distinctive brand on an indistinct sound.

Think about how stock libraries work. The most popular tracks rise to the top of search results. The top of the results gets licensed the most. The most licensed tracks show up in the most videos. That feedback loop means the "best" stock music is, almost by definition, the most overused. You're picking from the same shortlist as thousands of other people with the same deadline.

And audio sticks in memory in a way visuals often don't. A viewer might not remember the colour of an ad, but they'll recognise that ukulele-and-whistling track instantly. The trouble is, the recognition doesn't point to you. It points to every other brand that ever used it.

You would never use a competitor's logo. So why are you using their soundtrack?

What generic music costs you

• It flattens your emotional impact. Music written for nobody in particular moves nobody in particular.
• It blends you in exactly when you are trying to stand out.
• It caps your perceived value. Cheap-sounding is cheap-feeling.
• It builds no memory. Generic sound leaves no trace, so nothing accrues to your brand.

Each of these hits a different part of your marketing. Flat emotion means lower engagement on the video you paid to produce. Blending in means your ad gets scrolled past with the rest. A capped perceived value means you have to work harder to justify your price. And no memory means every campaign starts from zero, instead of stacking on the last one.

That last cost is the sneakiest. Brands that invest in a consistent sound get a compounding return: every ad, video and podcast intro reinforces the same audio cue. Brands that grab a new stock track every time get none of that. They pay for attention again and again and keep nothing. We dug into the commercial side of this in the sound of money.

The hidden licensing headaches

"Royalty-free" sounds like "free of problems." It isn't. It usually means you pay once, or pay a subscription, instead of paying ongoing royalties per use. The fine print still matters, and it trips up businesses all the time.

• Licence scope. Many licences separate personal, social, web, broadcast and paid advertising use. A track licensed for a YouTube video may not cover a TV spot or a paid social campaign.
• Subscription strings. Some libraries only cover content published while your subscription is active. Read the terms on what happens to past videos if you cancel.
• Content ID claims. Stock tracks registered with platform fingerprinting systems can trigger copyright claims on your own videos, even when you've licensed them properly. You then spend time disputing claims instead of running your campaign.
• No exclusivity. You can't stop anyone else, including a direct competitor, from licensing the exact same track tomorrow.
• No ownership. You can't trademark it, build a jingle from it, or turn it into an audio logo you control.

None of these mean stock music is illegal or dangerous. They mean it's rented, and rented things come with conditions.

When stock music is actually fine

We're not going to pretend every piece of content needs an original score. That's not realistic, and it's not smart spending.

Stock music works for internal videos, quick behind-the-scenes clips, low-stakes social posts, early testing of ad concepts, and background beds under speech where the music is barely noticed. In those cases, speed matters more than distinctiveness. Use it, pick carefully, and keep it quiet in the mix.

If you do use stock, pick like a pro. Skip past the first page of results, where the overused tracks live. Filter for less obvious moods and instrumentation. Avoid anything with a hook so memorable it announces itself. And keep a short internal list of approved tracks so your team isn't reaching for a different random song every week, which at least gives your everyday content some consistency.

The problem starts when stock becomes the sound of your flagship work: your brand film, your hero ad, your podcast theme, your product launch, your on-hold music, your trade show loop. These are the places people form their impression of who you are. That's where the soundtrack should belong to you.

Custom sound is an asset, not an expense

Music made specifically for your brand does something stock never can: it fits your message perfectly, it belongs to you, and it compounds. Every time people hear it, the association with your brand deepens instead of leaking away to a library. It turns sound from a disposable cost into an asset that gets more valuable the more you use it.

Custom music is also built to fit. A composer can hit the exact moment your product appears, swell when the offer lands, and resolve on your logo. Stock tracks are edited to fit your video after the fact, which is why so many ads have that awkward fade-out halfway through a musical phrase. When the music is written for the picture, the whole piece feels deliberate.

And one piece of original music can become a whole system. A full brand theme can be cut down to a thirty-second ad edit, a fifteen-second version, a short sting for social, a podcast intro, and a two-second audio logo. Same musical DNA everywhere, all instantly recognisable as you. That's the heart of sonic branding, and it's why the biggest brands treat their sound as seriously as their logo.

How to brief original brand music

Commissioning music doesn't have to be mysterious. A good brief gets you most of the way there. Here's the process we recommend:

1. Define the feeling in three words. Not genres. Feelings. "Confident, warm, restless" gives a composer far more to work with than "upbeat corporate."
2. Collect references, and say why. Pick a few tracks you love and note what you like: the energy, the instruments, the space. Also share what you hate, which is often just as useful.
3. List every place the music will live. Ads, social, website, events, podcast, phone line. This decides which edits and lengths you'll need.
4. Agree on ownership up front. Get clear on whether you're buying the music outright, licensing it exclusively, or licensing it for set uses and territories. Put it in writing.
5. Ask for stems and cutdowns. Separate instrument groups and multiple lengths make future edits fast and cheap.
6. Test it in context. Hear the draft under your actual video, not on its own. Music that sounds great alone can clash with a voiceover.

A worked example: one brand, two soundtracks

Say you run a Perth skincare brand. You spend serious money on a product shoot, beautiful macro footage, real texture, perfect light. Then you drop a popular stock track under the ad because the deadline is tomorrow.

The ad runs. Some viewers half-recognise the music from a mattress commercial. Others don't notice it at all. Six months later, you run a new campaign with a different stock track. Nothing connects the two in anyone's memory. You've paid twice and built nothing.

Now run it again with original music. You commission a short brand theme with a signature melody and a two-note audio logo. The launch ad uses the full theme. The social cutdowns use a fifteen-second edit. Your website hero video and your founder podcast use the same melody. By the second campaign, the first notes already feel like you before the logo appears. That's the difference between an expense and an asset.

What to do this week

• Audit your audio. List every piece of content with music in it and note where each track came from.
• Search your own tracks. Look up the stock tracks you use and see how many other brands are running them.
• Check your licences. Make sure each track is licensed for how you're actually using it, especially paid ads.
• Pick your flagship. Identify the one piece of content where original music would make the biggest difference and start there.

This matters more now, not less. With AI tools pumping out endless generic content, standing out is getting harder in every channel. We wrote about why distinctiveness is the only real defence in the AI content flood. Sound is one of the few places most of your competitors aren't even trying.

Sound as singular as you are

We create original music and sound design built around your brand, distinctive, emotional, and entirely yours. If you are investing in premium visuals and pairing them with music a thousand other brands are using, you are undercutting your own work. Our music and sound studio writes brand themes, ad scores, podcast music and audio logos that belong to you and nobody else. Let's make you sound as singular as you are.

Frequently asked questions

Is it okay to use royalty-free music in ads?

Yes, as long as the licence covers paid advertising in the places you're running it. Check the licence terms carefully, since some only cover personal or organic social use. Legally it's fine, but strategically, stock music in your main ads means competitors can use the same track and your brand builds no recognisable sound of its own.

How much does custom brand music cost compared to stock music?

Custom music costs more up front than a stock licence, and the price depends on length, complexity, the number of edits and how you own or license it. The better comparison is value over time. Stock music is a repeat expense that builds nothing. Original music is a one-off investment you can reuse across years of campaigns.

What does royalty-free actually mean?

Royalty-free means you pay once, or through a subscription, rather than paying ongoing royalties every time the music is used. It does not mean the music is free, and it doesn't mean you can use it anywhere. The licence still sets limits on things like advertising, broadcast and how long your rights last.

Can I own the music a composer makes for my brand?

Often, yes. Many composers offer a full buyout or an exclusive licence for brand work. The terms vary, so agree in writing who owns the composition and the recording, where and how you can use it, and whether the composer can reuse any of it. Settle this before work starts, not after the ad goes live.

So what makes a thumbnail stop the scroll? One clear subject, a readable human emotion, strong contrast against the feed, a question the viewer wants answered, and nothing that takes effort to decode. Get those five right and more people click the same video. Get them wrong and the quality of the video never gets a chance to matter.

The brain decides before it thinks

People are not carefully evaluating your thumbnail. They are flicking past hundreds of images while a fast, ancient part of the brain screens for one thing: is this worth stopping for? That decision happens before conscious thought, on instinct, emotion, and pattern. Which means a great thumbnail is not about looking nice. It is about triggering that instinct to stop.

Think about how you scroll. You are not reading titles one by one. You are pattern matching. Your eye jumps to faces, to bright shapes, to anything that looks different from what it just saw. Everything else blurs into wallpaper. The thumbnail that wins is the one that breaks the pattern and then, in the next beat, gives the viewer a reason to care.

That is a two-step job. Step one: be seen. Step two: be understood. Most thumbnails fail at step one because they blend in, or at step two because they are too busy to read at a glance.

Your content is not competing on quality. It is competing for a half-second of attention it has to win before anyone knows the quality exists.

What makes a thumb freeze

• A clear focal point. One dominant subject the eye locks onto instantly, clutter kills.
• Emotion. Human faces and strong feeling stop scrolls, the brain is wired to notice both.
• Contrast. It must pop against the feed around it, not blend into the noise.
• Curiosity. A visual that opens a loop the viewer needs to close by clicking.
• Instant readability. If it takes effort to understand, the moment is already gone.

Now let's turn each principle into something you can actually do on your next thumbnail.

One focal point, big

Pick the single most important thing in the frame and make it large. A face, a product, a before and after, a number. Crop tighter than feels comfortable. Remember most people will see your thumbnail on a phone, at the size of a postage stamp. Details that look lovely on your editing monitor simply disappear. If you have to explain what someone should look at, there are too many things in the frame.

Emotion you can read from across the room

A neutral face is almost as invisible as no face. Surprise, frustration, delight, disbelief, concentration: pick the emotion that matches the video and commit to it. It does not need to be cartoonish, but it needs to be readable in a split second. Eyes matter most. If the subject is looking at something in the frame, the viewer's eye follows.

Contrast with the feed, not just within the image

Contrast is relative. A bright red thumbnail stands out in a feed of muted tones and disappears in a feed of bright red. Look at what your competitors and your platform typically show, then go the other way. Separate your subject from the background with light, colour or a clean outline. Busy backgrounds are the fastest way to lose the focal point.

Curiosity, not confusion

The best thumbnails open a question. What happened here? What is in the box? Why is that number so low? The trick is that the viewer must understand enough to want the answer. A mystery with no context is just noise. A mystery with one clear clue is irresistible.

Readable in under a second

If you use text, keep it to a few words, in a bold, simple font, large enough to read on a phone. Never repeat the title word for word. The thumbnail and the title should work as a team: the image hooks, the title adds context, and together they make one clear promise.

Thumbnail, title and first seconds: one promise

A thumbnail does not work alone. It sits next to a title, and it hands the viewer to the opening seconds of the video. If those three pieces make different promises, you lose people at every handover.

The pattern that works: the thumbnail shows the most intriguing moment or outcome, the title frames the question, and the first few seconds of the video confirm the viewer clicked on the right thing. If the thumbnail shows a finished kitchen renovation, the video should open on it, not on three minutes of introduction. The same logic applies to ads, which is why the opening hook matters so much in the hook, story, offer formula.

Thumbnails that overpromise get clicks and then lose viewers immediately. That trains the platform, and your audience, that your content disappoints. Honest intrigue beats clickbait over any time frame longer than a day.

How to design a thumbnail that gets clicked, step by step

1. Write the promise first. In one sentence, what will the viewer get from this video? The thumbnail has to express that visually.
2. Pick the single most compelling moment. The reveal, the result, the reaction, the mistake. Plan to capture it as a still during the shoot, not as an afterthought.
3. Shoot the thumbnail on purpose. A frame grabbed from video is often soft or mid-blink. A dedicated photo with good light and a clear expression is sharper and more controllable.
4. Build it at thumbnail size. Design with the preview shrunk to phone size. If it does not read small, it does not work.
5. Add text only if it adds something. A few words that sharpen the curiosity, never a copy of the title.
6. Make two or three versions. Different expressions, crops or concepts. You will not know which works until the audience tells you.
7. Test and swap. Where a platform lets you test or change thumbnails after publishing, use it. Replace the underperformer and keep notes on what won.

Planning the thumbnail moment during the shoot is the big one. It is far easier to capture the right frame than to rescue a weak one in editing. It is also why a single well-planned shoot can supply a long run of content, stills included, as we explain in one video shoot, 90 days of content.

A worked example: the real estate walkthrough nobody clicked

Say you are a real estate agent on the Gold Coast posting property walkthrough videos. Your current thumbnail is a wide shot of the front of the house on a cloudy day, with the address in small white text across the bottom. It looks exactly like every other listing video in the feed. Nobody stops.

Apply the principles. Focal point: the best feature of the house is the view from the rooftop deck, so that becomes the image, cropped tight. Emotion: a still of you on the deck, genuinely taken aback by the view, looking out toward it so the viewer's eye follows. Contrast: the listings around you are grey and beige, so you shoot at golden hour and let the warm sky do the work. Curiosity: three words of text, "Wait for the roof", which gives one clue without giving away the reveal. Readability: large, bold, high contrast, no address in the thumbnail, because the title carries that.

Same house, same walkthrough, same agent. The new thumbnail gives someone a reason to stop and a reason to click. This is a hypothetical, but it is precisely the kind of change that shifts a video from invisible to watched.

The thumbnail mistakes that quietly kill your views

• Too much in the frame. Three people, a logo, a product, a headline and a background pattern. The eye has nowhere to land.
• Tiny text. If you need to zoom in to read it, nobody will.
• Brand templates that make every video look identical. Consistency is good. Sameness makes regular viewers think they have already seen it.
• Stock imagery. Viewers recognise generic stock instantly, and it signals generic content.
• Clickbait the video cannot back up. A cheap win on the click that costs you trust on every video after it.
• Ignoring dark mode and small screens. Check how it looks on a phone, in both light and dark app themes, before you publish.

How to measure whether a thumbnail is working

The number that matters most is click-through rate: of the people who were shown the thumbnail, how many clicked. Most video platforms report it, often alongside impressions. A higher click-through rate on similar impressions means your gate is letting more people in.

But never read it alone. Pair it with how long people watch after clicking. A thumbnail that lifts clicks and tanks watch time is overpromising. A thumbnail that lifts both is doing its job. Compare against your own previous videos on similar topics, not against someone else's channel, and give each test enough impressions to mean something before you call it.

The highest-leverage image you own

Think about the math. The thumbnail is one image, but it decides the fate of everything behind it. Improving it is one of the highest-return moves in all of content, the same video, the same effort, suddenly seen by multiples more people because the gatekeeper finally let them in.

We design thumbnails and visual hooks engineered around how attention actually works, so your best content stops being invisible and starts getting seen. Through our photo and video production work at LIQID NOISE, we plan the thumbnail moment before the camera rolls, shoot it properly, and build versions worth testing. If your videos are underperforming, look at the gate before you blame the room. Let's make yours impossible to scroll past.

Frequently asked questions

What makes a good video thumbnail?

A good thumbnail has one clear focal point, a readable human emotion or striking subject, strong contrast with the feed around it, and a hint of curiosity that makes the viewer want to click. It must be understandable in under a second at phone size. Any text should be a few bold words that add to the title, not repeat it.

Should I put my face on my YouTube thumbnails?

Often, yes, if the expression is clear and matches the video. People instinctively notice faces and emotion, which helps a thumbnail stand out in a busy feed. A neutral or blurry face adds little, though. If the video is about a product, place or result, that subject may be the stronger focal point. Test both approaches on your own audience.

How much text should a thumbnail have?

As little as possible. A few large, bold words usually work better than a sentence, because most people see thumbnails on a phone at a very small size. Use text to add a clue or sharpen the curiosity, and let the title carry the full context. If the text has to be read slowly, it is too long.

How do I know if my thumbnail is the problem?

Check click-through rate against impressions. If a video is being shown to plenty of people but few click, the thumbnail and title are likely the problem. If people click but leave quickly, the problem is more likely a mismatch between the promise and the opening seconds. Compare against your own past videos on similar topics.

Here is what they do not realise: you do not have to become a writer, and it does not have to take a year. With the right system, a busy founder can go from expertise to a finished book in about 90 days.

How? You stop trying to write it alone. You map the book first, talk your expertise out in guided interviews, let skilled writers shape those words into chapters that still sound like you, and review in tight rounds. Your job becomes a few focused hours a week, not a second career.

Your knowledge is already there

You have already done the hard part. Years of doing the work, solving the problems, learning the lessons, the book already exists inside you. The bottleneck was never the ideas. It was the process of getting them out of your head and onto the page without clearing your calendar for twelve months. Solve the process, and the book follows fast.

Think about how much you already explain every week. The same questions from prospects. The same lessons you teach new staff. The same stories you tell on sales calls because they land every time. That's not random talk. That's chapter material, repeated and refined over years. Most founders have tested their book's best ideas hundreds of times without ever writing them down.

You do not have a knowledge problem. You have an extraction problem. The book is already in your head, it just needs a system to pull it out.

Why founders stall on their book

It helps to name the traps, because nearly every stalled book falls into one of them.

• Starting at page one. Opening a blank document and writing "Chapter 1" with no map. Three weeks later you've rewritten the introduction six times and have no idea what chapter four is about.
• Trying to write like an author. Founders who are brilliant out loud freeze when they try to sound "literary." The result is stiff, and it doesn't sound like them.
• Writing in leftover time. The book gets whatever's left after the business takes its share. Which, in a growing business, is nothing.
• No deadline, no editor. Nobody is waiting for chapter three, so chapter three never arrives.
• Endless polishing. Treating every draft as the final draft, so nothing ever gets finished enough to improve.

Every one of those is a process problem. None of them has anything to do with whether you're smart enough or interesting enough to have a book.

How the 90-day book works

• Structure first. Map the whole book before writing a word, so there is no staring at blank pages.
• Talk, do not type. Your expertise gets captured through guided conversation, not lonely writing sessions.
• Ghost-shaped drafting. Skilled writers turn your words into polished chapters that still sound like you.
• Tight feedback loops so it stays on message without endless rewrites.
• Publish with intent, professionally edited, designed, and released to actually build your authority.

The reason this works is division of labour. You do the part only you can do: know things, have opinions, tell stories from real experience. Writers do the part they're trained for: structure, flow, clarity. Editors and designers do theirs. Nobody is asked to be good at everything, which is exactly why the book actually gets finished.

The 90 days, phase by phase

Here's roughly how the time breaks down. The exact timing flexes with the book, but the shape holds.

1. Weeks 1 to 2: strategy and structure. Who is the book for? What should a reader believe, feel and do after finishing it? What does it need to do for your business: win clients, open speaking doors, support a launch? From there, the chapters get mapped with key points, stories and examples for each. This outline is the single most important document in the whole project.
2. Weeks 3 to 6: extraction. A series of recorded interviews, usually an hour or so each, working through the outline chapter by chapter. You talk. A good interviewer pushes for the specific example, the exact mistake, the story behind the principle. This is where your voice gets captured.
3. Weeks 5 to 9: drafting. Writers turn the transcripts into chapters, keeping your phrasing, your stories and your opinions intact. Drafting starts while the later interviews are still happening, which is how the timeline stays tight.
4. Weeks 8 to 11: review rounds. You read chapters in batches and mark what's wrong, missing or not quite you. Tight, structured feedback beats vague "I'm not sure about this bit" notes. Two focused rounds usually do more than five loose ones.
5. Weeks 11 to 13: edit, design and production. A professional edit, then cover and interior design, then files ready for print and ebook. Your release plan gets locked in alongside.

Your time across all of that is mostly interviews and reading. For most founders that's a few hours a week, which fits around a business in a way that "write a book" never does.

A worked example: the consultant with the book in her head

Say you run a small HR consultancy in Sydney. For eight years you've helped growing businesses hire their first managers. You have a way of doing it, a set of mistakes you see over and over, and stories that make clients laugh and wince in the same breath. You've wanted a book for five of those eight years.

Now picture the 90 days. In the first fortnight you agree the book is for founders hiring their first manager, and its job is to bring in consulting clients. The outline has ten chapters, each built around a mistake you've watched businesses make.

For the next month you do one interview a week, sometimes two. You tell the story of the founder who promoted her best salesperson and lost both a salesperson and a manager. You explain the three questions you ask in every reference check. The writer asks "what happened next?" until each story has a proper ending.

Chapters start arriving in week five. The first one is close but a bit polite. You say so. The writer adjusts, and by chapter three it sounds like you on a good day. By week thirteen you have an edited, designed book with a clear call to work with you at the end of every chapter.

You didn't write a word from scratch. You did what you already do every day: explain your thinking out loud to someone who wants to learn it.

What makes a founder's book actually work

Finishing is the first win. A book that does its job for your business is the second, and it's decided long before the cover is printed.

• A narrow reader. "Business owners" is too broad. "Founders hiring their first manager" is a book someone recognises as written for them.
• A clear promise. The reader should know from the title and back cover exactly what they'll be able to do after reading it.
• Stories, not just frameworks. Frameworks get nodded at. Stories get remembered and retold. We go deeper on this in how to write a business book people actually finish.
• A path to you. Each chapter should make it natural for the right reader to want your help. That's the logic behind a book that closes deals while you sleep.
• Professional finish. A proper edit and a proper cover. Readers judge the book, and you, on both.

Publishing and launching without losing momentum

The manuscript is done. Now the question is how it reaches readers. For most founders on a 90-day timeline, self-publishing or hybrid publishing is the practical route, because it keeps control and speed in your hands. Traditional publishing can suit some authors, but it usually adds a long wait before release. If you're weighing that decision, read self-publishing vs. traditional first.

Then there's the launch, which is where many good books quietly disappear. A launch doesn't need to be huge. It needs to be deliberate: your email list, your clients, your network, a handful of podcast or event appearances, and a clear plan for how the book gets into the hands of the people you want to work with. The book isn't the finish line. It's a sales and authority asset you'll use for years.

Done beats perfect, and done beats someday

The book that gets written imperfectly this quarter beats the perfect book you will "definitely" start next year. A finished, published book working for your business in 90 days is worth infinitely more than a flawless manuscript that never exists. The only bad book is the one you never wrote.

Perfectionism feels like high standards. Mostly it's fear wearing a nice jacket. A finished book can be revised, updated and released as a second edition. An unwritten book can't be improved at all.

Your book, without the lost year

We run the entire process, extracting your expertise, shaping it into a compelling book, and publishing it, without swallowing your calendar. If a book has been on your "someday" list for years, the problem was never you. It was the process. Let's give you one that works.

That's what our books and publishing team exists to do: strategy, interviews, drafting, editing, design and a launch plan, all in one place. Bring the knowledge. We'll bring the system that gets it onto the page and into your readers' hands.

Frequently asked questions

Can you really write a book in 90 days?

Yes, if you use a system instead of writing alone. A 90-day book relies on mapping the structure first, capturing your expertise through recorded interviews and having skilled writers draft chapters while interviews continue. Review happens in tight rounds, then editing and design follow. The founder's time is mostly talking and reading, which is why it fits around a busy business.

Is it cheating to use a ghostwriter for a business book?

No. The ideas, stories, opinions and expertise are yours. A ghostwriter's job is to shape your words into clear chapters, the same way an editor or speechwriter helps a leader communicate. Many business books are written this way. What matters is that the book genuinely reflects your thinking and sounds like you, which is exactly what interview-led drafting is designed to protect.

How much time does a founder need to spend on their book?

With an interview-led process, most of a founder's time goes into recorded conversations early on and reading draft chapters later. For most people that works out to a few hours a week over about three months. It's far less than writing from scratch, because you're not staring at a blank page or rewriting the same chapter over and over.

Should a business book be self-published?

For most founders, self-publishing or hybrid publishing makes sense because it's faster and keeps you in control of content, pricing and how the book is used in your business. Traditional publishing can bring bookstore distribution and prestige but usually involves a longer timeline and less control. The right choice depends on your goals for the book and how quickly you need it.

That is what makes speed the most dangerous problem in your business: it is invisible, and it is bleeding you right now.

Does website speed really affect sales? Yes. A slow site loses visitors before they see your offer, converts fewer of the ones who stay, and ranks lower in search, so it costs you twice. The good news is that speed is one of the most fixable problems on any website, and every improvement recovers customers you've already paid to attract.

The numbers are merciless

Study after study says the same thing. Google's Need for Mobile Speed research found 53% of mobile visitors leave a page that takes longer than three seconds to load. Portent's analysis of more than 100 million page views found conversion rates fall as load time climbs. On mobile, where most of your audience actually is, it is worse, because connections are slower and patience is thinner.

Think about how you browse. You tap a link on your phone, standing in a queue or sitting on the couch with half your attention elsewhere. The screen stays white. A spinner turns. You hit back and tap the next result. You didn't make a conscious decision to reject that business. You just moved on. Your customers do exactly the same thing to you.

And the damage isn't limited to people who leave. Visitors who stick around on a sluggish site are already irritated. Buttons that lag, layouts that jump as images load, forms that freeze when you tap submit. Each one chips away at trust right at the moment you need it most.

Speed is not a technical detail. It is the first promise your brand makes, and a slow site breaks it before you say a word.

Why speed also decides your rankings

Google has made it plain that its ranking systems reward a good page experience, and speed is a big part of that. Which means a sluggish site does not just lose the visitors who arrive, it quietly loses the visibility that would have brought more of them. Slow is a tax you pay twice: once in rankings, again in conversions.

Google measures this with Core Web Vitals, a set of real-world metrics covering three things a visitor actually feels:

• Loading. How quickly the main content appears. Largest Contentful Paint (LCP) tracks when the biggest visible element, usually your hero image or headline, finishes rendering.
• Responsiveness. How quickly the page reacts when someone taps or clicks. Interaction to Next Paint (INP) measures that delay.
• Visual stability. Whether things jump around as the page loads. Cumulative Layout Shift (CLS) catches the moment you go to tap a button and an ad pushes it down the screen.

These are measured from real visitors on real devices, not just a lab test on a fast office connection. That matters, because your site might feel fine on your laptop and still fail for the people who actually buy from you. We go deeper into the current rules in Core Web Vitals in 2026.

What actually makes a website slow

• Bloated images and heavy scripts that make phones crawl.
• Cheap, overloaded hosting shared with a thousand other sites.
• Template code stuffed with features you never use but every visitor still downloads.
• No caching, no optimisation, no discipline, just "it works on my laptop."

A few more that show up constantly:

• Plugin pile-ups. Every plugin added to fix a small problem adds its own code, styles and requests. Twenty of them and the site is carrying a backpack full of bricks.
• Third-party scripts. Chat widgets, tracking pixels, review badges, social embeds and pop-up tools. Each one loads from someone else's server, on someone else's schedule.
• Autoplaying video and sliders. Big, heavy and usually ignored. The classic homepage carousel costs speed and rarely earns a click.
• Custom fonts loaded badly. Several font families and weights can hold up the text people came to read.

The fix is not magic. It is engineering: lean code, optimised assets, modern architecture, and a relentless focus on the moment a real person on a real phone taps your link.

How to test your site's speed today

You don't need to guess. You can check where you stand in about ten minutes:

1. Run Google PageSpeed Insights on your homepage and your most important money page, such as your main service page or top product. Look at the mobile result first.
2. Check the real-user data. If your site has enough traffic, the report shows how actual visitors experienced it. That matters more than the lab score.
3. Open Search Console's Core Web Vitals report. It groups your pages into good, needs improvement and poor, so you can see whether the problem is one template or the whole site.
4. Test on a real phone on mobile data. Not office wifi. Tap your own ad or search result and watch what a customer sees.
5. Note the biggest offenders. The reports list the specific images, scripts and resources slowing things down. That's your fix list.

Don't obsess over hitting a perfect score. The goal is a site that feels instant to real people and passes the metrics that matter, not a trophy.

The fixes, in order of impact

Most sites can get dramatically faster by working through this list from the top:

1. Compress and resize images. Serve modern formats like WebP or AVIF, at the size they're actually displayed. A phone doesn't need a photo sized for a billboard. This alone is often the biggest win.
2. Cut third-party scripts. Audit every widget and tag. Remove what nobody uses, and load the rest after the main content so it doesn't block the page.
3. Lazy-load what's below the fold. Images and videos further down the page should load as people scroll, not before the visitor sees anything.
4. Upgrade hosting and add a CDN. Good hosting and a content delivery network serve your site from locations close to your visitors, which matters when your customers are spread across Australia or the world.
5. Turn on caching. Stop rebuilding the same page from scratch for every visitor.
6. Trim the code. Remove unused plugins, styles and scripts. If a bloated theme is the root cause, a rebuild may be cheaper than endless patching.
7. Reserve space for images and embeds. Set dimensions so the layout doesn't jump as things load.

Some of these are quick wins you can do in an afternoon. Others need a developer. And sometimes the honest answer is that the site was built cheaply on a foundation that can't be made fast. We've written about the real cost of a cheap website, and speed is usually the first place it shows.

A worked example: the leak nobody could see

Here's a hypothetical. Say you run a landscaping business on the Central Coast. You spend $2,000 a month on Google Ads, sending people to a quote page built on a heavy template. The page has a full-screen video background, a slider, a chat widget, three tracking scripts and a quote form that loads last.

On office wifi it seems fine. On a phone in a car park on patchy mobile data, the form takes long enough to appear that plenty of people give up. You don't see them leave. You just see "clicks" in your ad account and not enough quote requests, so you assume the ads aren't working and raise the budget.

Now you fix the page. The video becomes a single compressed photo. The slider goes. The chat widget loads after the form. Images are resized and the tracking is cleaned up. Nothing about the offer changed, and the ad spend stays the same, but more of the people you're already paying for now reach the form and use it.

The figures here are illustrative, but the logic is exactly why speed is such a strong lever. You didn't buy more traffic. You stopped wasting the traffic you already had. If your landing pages need more than speed, our guide to landing pages that print money covers the rest.

The cheapest growth lever you own

Most growth costs money, more ads, more content, more outreach. Speed is different. You already paid for the traffic. Making your site faster does not buy you more visitors, it stops you from losing the ones you already have. That is pure profit, recovered.

And the benefit stacks across every channel. Your ads convert better because visitors reach the offer. Your SEO improves because search engines reward the experience. Your email clicks land on pages that load. Every dollar you spend on marketing works harder when the destination is fast.

To prove it to yourself, measure before and after. Record your mobile Core Web Vitals, bounce rate and conversion rate on your key pages, make the fixes, then compare the same numbers a month later. That turns speed from a vague technical chore into a line item with a return you can see.

Common speed mistakes

• Testing on your own device. Your fast laptop and office connection hide the problem your customers see.
• Chasing a score instead of the experience. A perfect lab score means little if real visitors still wait.
• Adding a plugin to fix a plugin. Speed plugins stacked on a bloated site treat symptoms, not the cause.
• Fixing it once and forgetting it. Every new widget, tracking tag and oversized image slowly drags speed back down. Check it every quarter.
• Only fixing the homepage. Your service pages, product pages and landing pages are where the money is made. Test those too.

Fix the leak you cannot see

We build sites that feel instant, because a fast site is a site that respects the customer, ranks higher, and converts more. If yours makes people wait, it is making them leave. Let's fix the leak you cannot see. Our web development team builds lean, fast sites and rescues slow ones.

Frequently asked questions

How fast should my website load?

Fast enough that the main content appears almost immediately on a mobile phone, and the page responds instantly when someone taps. The practical benchmark is passing Google's Core Web Vitals for real users, which you can check in PageSpeed Insights or Search Console. If your pages pass for mobile visitors and feel instant on a real phone, you're in good shape.

Does website speed affect SEO?

Yes. Google uses page experience signals, including Core Web Vitals, as part of how it ranks pages. Speed won't outrank better content on its own, but when pages are otherwise comparable a faster, smoother experience helps. A slow site also loses visitors who do arrive, which weakens the engagement and conversions that make a site worth ranking.

What is the most common cause of a slow website?

Oversized images are one of the most common culprits, along with too many third-party scripts and plugins, heavy page builder themes and cheap shared hosting. Running a PageSpeed Insights test shows exactly which files slow your pages down, so you can fix the biggest offenders first instead of guessing.

Is it cheaper to speed up my website or rebuild it?

It depends on the foundation. If the site is reasonably built and the problem is images, scripts and hosting, optimising it is usually cheaper. If the theme itself is bloated or the site relies on dozens of plugins, patching can cost more over time than a lean rebuild. A proper audit tells you which situation you're in.

Vanity metrics are the sugar of marketing: an instant hit that leaves you worse off. They make you feel productive while quietly steering every decision away from the only thing that keeps the lights on, money in the door.

Here is the plain version. A vanity metric is any number that can rise while your revenue stays flat: likes, followers, impressions, views, reach. They are not useless, but they are not the goal. The numbers that run a business are what a customer costs you, what a customer is worth, how many visitors act, and which channels actually return money. Track those first and let the applause take care of itself.

The seductive lie of the like

Here is the trap. Content that gets the most likes is often content that sells the least, because "entertaining" and "persuasive" are different jobs. So you chase engagement, get rewarded with applause, and slowly train your entire marketing to optimise for a number that never once paid an invoice.

It happens gradually. A funny behind-the-scenes clip does great numbers. The next week you make another one. The sales-focused post explaining your offer gets a quarter of the likes, so you post fewer of those. Six months later your feed is a comedy channel with a logo on it, the follower count is up, and the enquiry form is quiet. Nobody decided to stop selling. The metric decided for you.

Platforms make this worse because the vanity numbers are the ones they show you first. They are big, they update in real time, and they feel good. The numbers that matter are buried in other reports, or live in your bank account and CRM, where the social dashboard never looks.

If a metric goes up and your bank balance does not, you are not marketing. You are performing.

Vanity metrics vs money metrics

The simplest test for any number in a report: if this went up tomorrow and nothing else changed, would the business be better off? If the answer is "not necessarily", it is a vanity metric.

• Followers. Could be buyers. Could be bots, competitors and people who liked one meme. On its own it says nothing about revenue.
• Likes and reactions. A measure of approval, not intent. People happily like content from businesses they will never buy from.
• Impressions and reach. How many times something was shown. Useful for diagnosing delivery, meaningless as a goal.
• Website traffic. Better, but still a vanity number unless you know what those visitors do. Fewer, better visitors often make more money than a flood of the wrong ones.

That last point is becoming more important every month. Many businesses are seeing traffic fall while enquiries and revenue hold steady or rise, because the visitors who remain are further along the buying journey. If you judge that on a traffic graph, it looks like failure. We look at this shift closely in fewer people are searching, the ones left are ready to buy.

The metrics that actually run a business

• Cost to acquire a customer, what you pay to turn a stranger into a buyer.
• Customer lifetime value, what that buyer is worth over the whole relationship.
• Conversion rate, how many people who arrive actually act.
• Return on ad spend, dollars out for every dollar in.
• Revenue per channel, which efforts actually pay, and which just feel good.

Notice none of those are likes. Track these and something powerful happens: you can finally tell the difference between marketing that works and marketing that merely flatters you.

How to calculate the numbers that matter

None of this needs expensive software. A spreadsheet and an honest hour a month will do. Here is how to work each one out.

1. Customer acquisition cost. Total marketing and sales spend for a period, divided by the number of new customers in that period. Include ad spend, agency fees, software and the value of your own time if you want the real figure.
2. Customer lifetime value. Average purchase value, multiplied by how many times an average customer buys, multiplied by your gross margin. Use your actual history, not hopes.
3. The ratio between them. Lifetime value divided by acquisition cost. If a customer is worth less than they cost to win, more marketing only loses money faster.
4. Conversion rate. Enquiries or sales divided by visitors, for each key page or campaign. Track it separately for each channel, because an average hides the winners.
5. Return on ad spend. Revenue attributed to ads divided by ad spend. Then sanity check it against actual sales, because platforms tend to be generous when counting their own results.
6. Revenue per channel. Ask every new customer how they found you and record it. Combine that with your analytics. It will not be perfect, and it will still be far better than guessing.

The single most useful habit on that list is the last one. Asking "how did you hear about us?" costs nothing, and it regularly reveals that a channel nobody thought about is quietly bringing in the best customers.

A worked example: the gym that was winning on Instagram

Say you own a boutique gym in Adelaide. Your Instagram is thriving: trending audio, workout fails, member shout-outs. Followers are climbing every month and you are spending $1,500 a month on a content creator and a small ad budget to keep it going. Meanwhile, you also run a simple referral offer where members get a free month when a friend joins. It costs you almost nothing to run and nobody thinks of it as marketing.

You finally sit down and ask every member who joined in the last six months how they found you. Imagine the answer comes back like this: most new members came from friends and the referral offer, a handful came from Google searches for gyms nearby, and only a couple mention Instagram at all. Your most visible channel turns out to be your weakest salesperson, and your invisible one is doing the heavy lifting.

That does not mean you delete the Instagram account. It means you stop judging it on followers, redirect some of that budget to making the referral offer louder, tidy up your Google Business Profile, and change the Instagram content to feature real member results with a clear call to book a trial. Then you measure trial bookings, not likes. This is a hypothetical, but the pattern of a loud channel hiding a quiet one is extremely common.

Where vanity metrics still earn their place

This is not a war on engagement. Some top-of-funnel numbers are useful, as long as you use them as diagnostics rather than goals.

• Reach and impressions tell you whether your content or ads are being delivered at all. If they drop sharply, something is wrong with delivery.
• Engagement rate can signal which messages resonate, which is useful input for ads and sales pages.
• Video watch time shows whether your hooks and stories hold attention.
• Email and SMS list growth is closer to a real asset, because you own the audience. That is one reason email remains the highest-ROI channel nobody respects.

The rule is simple. Use these to explain why the money metrics moved. Never use them to replace the money metrics.

Common measurement mistakes to fix first

• Trusting platform reporting blindly. Each ad platform counts its own contribution generously. Compare their numbers to your real sales.
• No tracking on the website. If form submissions, calls and purchases are not recorded as conversions, you are guessing. Start there.
• Judging too early. Some channels take weeks or months to pay back. Set a review window before you start.
• Averaging everything together. A blended cost per lead hides the channel that is losing money and the one that deserves double.
• Reporting to impress instead of to decide. If a report never changes a decision, it is decoration.

If your ad account is a big part of spend, it is worth running a proper audit on where the money is going. Our guide on the audit that finds your wasted ad spend walks through it.

Build a one-page dashboard you will actually read

The best marketing dashboard fits on one page and takes five minutes to update. Put the money metrics across the top: new customers, acquisition cost, revenue by channel and the lifetime value to cost ratio. Underneath, list each channel with what you spent, what it produced and whether it is trending up or down against last month.

Keep the diagnostic numbers, like reach and engagement, on a second page for whoever runs that channel. The owner and the leadership meeting only need page one. Review it monthly, and make one decision every time you do: cut something, scale something, or test something. A dashboard that never triggers a decision is just another vanity metric.

What changes when you count what counts

You stop pouring effort into content that entertains but does not convert. You double down on the quiet channels that actually sell. You make decisions with numbers instead of vibes. And your marketing starts behaving like an investment with a return, not a hobby with an audience.

Start this week with three things. Add "how did you hear about us?" to every enquiry form and every first call. Work out your rough acquisition cost for last quarter. Then look at your monthly marketing report and cross out every number that could rise without revenue rising with it. Whatever is left is your real dashboard.

We build marketing on a foundation of real metrics, tracking every dollar from click to customer, so you always know what to cut and where to scale. That is the core of our digital marketing work at LIQID NOISE. If you cannot say what a customer costs you, you are flying blind. Let's put instruments in the cockpit.

Frequently asked questions

What are vanity metrics in marketing?

Vanity metrics are numbers that look impressive but do not reliably connect to revenue, such as likes, followers, impressions, views and raw traffic. They can rise while sales stay flat. They are useful for diagnosing delivery and resonance, but they should not be the goal of your marketing or the main thing you report to decide where money goes.

Which marketing metrics matter most for a small business?

Start with customer acquisition cost, customer lifetime value, conversion rate, return on ad spend and revenue per channel. Together they tell you what a customer costs, what they are worth, whether visitors act and which efforts pay. You can calculate all of them with a spreadsheet, your sales records and a simple habit of asking new customers how they found you.

Are followers worth anything for a business?

Followers have some value as a warm audience you can reach, but the number alone says little. A small following of likely buyers can outsell a huge following of casual scrollers. Judge social media on the enquiries, trials and sales it produces, and treat follower growth as a side effect of content that attracts the right people, not the target.

How do I track which marketing channel brings in customers?

Combine two sources. Set up conversion tracking on your website so enquiries, calls and purchases are recorded by source. Then ask every new customer how they found you and log the answer. Analytics catches the digital path, and the question catches word of mouth, offline and AI-assistant recommendations that tracking tools often miss.

Plenty of high-returning ad accounts run on modest daily budgets. Plenty of poor performers burn tens of thousands a month. The difference is almost never the money. It is the match.

So can a $100-a-day budget really beat a $10k agency retainer? Yes, when the small budget has nailed message-to-market match and the big one hasn't. The right message, in front of the right people, with an offer they want, makes every dollar work. The wrong message just loses money faster, however much you pour in.

Message-to-market match beats budget

When the right message meets the right person with the right offer, magic happens, a small budget compounds into a flood of customers because every dollar is working. When the message is vague, the audience is wrong, or the offer is weak, more money just means losing faster. You cannot outspend a bad message. You can only afford to fail bigger.

Break "match" into its three parts and it gets practical.

• Market: Who, specifically, has the problem you solve and the money to fix it? "Homeowners" is not a market. "Owners of 1970s brick homes in Canberra who want better insulation before winter" is.
• Message: What do those people already believe, fear and want, in their own words? The ad that wins sounds like the thought already in their head.
• Offer: What exactly do they get if they respond, and why is it a better deal than doing nothing? A weak offer can sink a perfect message. We break down how to build one in the Godfather offer.

Get all three aligned and the ad platforms do the rest. Modern delivery systems are very good at finding more people like the ones who respond. They are terrible at rescuing an ad nobody wants to respond to.

A big budget on a weak message is just an expensive way to find out your message is weak.

Why lean often beats bloated

• Focus. A tight budget forces you to nail one audience and one offer instead of spraying cash everywhere.
• Speed. Small, sharp tests reveal what works fast, so you double down before big spenders finish their first report.
• Discipline. When every dollar counts, you obsess over the creative and the offer, the things that actually move numbers.
• No agency drag. No bloated retainers, no layers, no account manager who has never seen your customer.

There's a quieter reason too. Big budgets hide problems. When an account spends heavily across a dozen campaigns, a bad ad can keep running for weeks because it's a small line in a big report. On $100 a day, a bad ad sticks out by Wednesday. Constraint gives you clarity. Clarity gives you better decisions. Better decisions compound.

Where the money actually leaks in big accounts

If you've ever paid a large agency and felt the results didn't match the invoice, it usually comes down to a few familiar leaks.

• Structure for the report, not the result. Dozens of campaigns and audiences that look impressive in a monthly deck and split the budget so thin that nothing gets enough data to learn.
• Creative made once and left to rot. A shoot at the start of the quarter, then the same ads running long after the audience has stopped noticing them.
• Optimising for the wrong event. Chasing cheap clicks or cheap form fills instead of customers who actually pay.
• Distance from the customer. Account managers juggling many clients, writing ads for a business they've never seen from the inside.

None of that is about money. It's about attention. We walk through how to find this kind of waste in your own account in the audit that finds your wasted ad spend.

Spend earns the right to scale

Here is the real sequence the big spenders skip: you do not scale to find what works, you find what works, then scale. A lean budget that has cracked message-to-market match is a coiled spring. Add fuel to a fire that is already burning and it roars. Add fuel to a message that is not working and you just make a bigger mess.

In practice, "earning the right" means hitting a few clear conditions before you raise spend: a cost per customer you can live with, consistent results over a couple of weeks rather than a lucky few days, and enough profit per sale that more volume means more money in the bank. Then you increase budget in steps, watching whether results hold, instead of multiplying it overnight and hoping.

How to run a $100-a-day test that actually teaches you something

A small budget only wins if it's run with intent. Here's the sequence.

1. Define one goal. One product or service, one type of customer, one action you want them to take. Not "brand awareness plus leads plus sales."
2. Know your numbers first. What's a customer worth to you? What can you afford to pay to get one and still make money? Without this, you can't tell a good result from a bad one.
3. Write several distinct angles. Not five versions of the same ad with a different button colour. Genuinely different reasons to buy: save time, avoid a costly mistake, look better, feel safer. The hook, story, offer formula is a solid frame for each.
4. Keep the structure simple. One campaign, a broad audience in the right location, a handful of ads. Let the platform find the people. Your job is to give it good creative to work with.
5. Track the right event. Make sure the platform is told when a real lead or sale happens, not just when someone lands on a page.
6. Give it time. Judge over a week or two, not a day. Small budgets need patience to gather enough signal.
7. Cut losers, feed winners, replace often. Turn off ads that clearly aren't working, keep the ones that are, and add fresh variations of the winning angle so the account never goes stale.

A worked example: the physio who stopped boosting posts

Say you run a single physio clinic in Hobart. You've been boosting Facebook posts for a while, and before that you paid an agency a monthly retainer plus ad spend. Results were fine on paper, lots of reach and likes, but your appointment book didn't feel it.

Now picture the lean version. You set a budget of $100 a day. You pick one service: treatment for running injuries. You pick one customer: local runners who are sore and worried about missing their next event. You write four angles. One about getting back to running faster. One about the mistake of resting and hoping. One about a named local running route with a sore knee at the end of it. One simple offer: a first assessment with a clear plan.

You film the ads on a phone in the clinic, with the physio talking straight to camera. You send bookings back to the platform as conversions. After two weeks, one angle is clearly booking appointments at a cost you're happy with and the others aren't. You pause the losers, film three new versions of the winner and keep going. Only when bookings hold steady for a few weeks do you step the budget up.

That's the whole trick. Nothing clever. Just a narrow target, a message that sounds like the customer, a real offer, honest tracking and the discipline to scale only what's proven.

The mistakes that make small budgets lose

To be clear, small budgets don't win automatically. They lose too, usually for these reasons.

• Spreading $100 across ten audiences. Each gets $10 and nothing learns anything.
• Changing everything every day. Constant tinkering resets the platform's learning and makes results impossible to read.
• Blaming targeting when it's the creative. Most underperforming accounts don't have an audience problem. They have an ad problem. We make that case in your ROAS problem is a creative problem.
• Sending traffic to a weak page. A great ad pointing at a slow, confusing page wastes the click.
• Scaling on a lucky week. Doubling budget after three good days is how lean accounts turn into bloated ones.

When a bigger budget does make sense

None of this means budget is irrelevant. Once you've found a message that works, more money genuinely means more customers, up to the point where you run out of people who want what you sell. Bigger markets, higher-value products and proven offers can absorb much larger spend profitably.

The point is order. Budget amplifies whatever you give it. Give it a proven message and it amplifies profit. Give it a guess and it amplifies the guess. The businesses that end up spending big and winning almost always started small and got the match right first.

There's also a practical ceiling worth knowing about. A local business can only serve so many customers in a week. If your clinic, workshop or kitchen is already booked out, extra ad spend doesn't buy growth. It buys frustrated leads you can't serve. Sometimes the smartest budget decision is to hold spend steady, raise your prices, or fix capacity first, then scale again once you can actually deliver on the demand you're paying to create.

Stop outspending, start outmatching

We build campaigns that earn the right to scale, nailing the match first on a lean budget, then pouring on fuel once the numbers prove themselves. If you are outspending your competitors and still losing, the problem is not your budget. Let's find what is actually holding you back.

That's the approach behind our advertising work: close to the customer, obsessive about creative and offer, and honest about what the numbers say. Whether you're spending $100 a day or much more, the question is the same. Is every dollar working?

Frequently asked questions

Is $100 a day enough for Facebook or Google ads?

For many local and small businesses, yes. $100 a day is enough to test a few distinct ad angles against one clear goal and learn what works within a couple of weeks. The key is focus: one offer, one type of customer, a simple campaign structure and proper conversion tracking. Spread that same budget across many audiences and campaigns and it stops being enough.

Why is my agency spending a lot on ads with poor results?

Usually because the message, audience and offer don't match, and extra spend can't fix that. Common causes include stale creative, campaigns split too thin to learn, tracking cheap clicks instead of real customers, and account managers too far from your business. Ask what the actual cost per paying customer is, and whether the ads have changed recently.

When should I increase my ad budget?

Increase spend once your ads consistently bring in customers at a cost you can afford over a period of weeks, not days. Raise the budget in steps and watch whether cost per customer holds. If results slip sharply, pull back and refresh the creative before trying again. Scaling a proven message works. Scaling a guess just loses money faster.

What is message-to-market match in advertising?

Message-to-market match means your ad speaks to a clearly defined group of people, in language that reflects what they already think and want, with an offer that suits them. When all three line up, ads convert efficiently even on small budgets. When one is off, such as a vague audience or weak offer, results suffer however much you spend.

So should you charge more? If you deliver real quality, almost certainly yes. Premium pricing works because buyers use price as evidence of value, and a higher price, backed by positioning and an experience that matches it, attracts better customers, funds better work and makes you the serious choice instead of the cheap one.

The catch is the word "backed." Raising prices without changing anything else is a gamble. Raising prices as part of a deliberate move up-market is a strategy. This article is about the second one.

Price is a message

Before anyone experiences your product, your price has already told them a story. Low price whispers "ordinary, replaceable, risky." High price signals "premium, proven, worth it." We are wired to assume expensive means better, and when you compete on cheap, you are actively telling the market to expect less of you.

Think about how you shop for something you can't easily judge. A surgeon. A lawyer. A bottle of wine for someone you want to impress. You don't have the expertise to assess quality directly, so you reach for the nearest available signal. Price is often that signal.

Your customers are doing exactly the same thing with you. Most of them can't evaluate your work in advance. They're guessing. And when you're the cheapest option on the table, you've handed them a very clear clue about how to guess.

When you cut your price, you are not just earning less. You are announcing that you are worth less.

The trap of cheap

• Thin margins leave nothing to reinvest in quality, service, or growth.
• Bargain hunters are the most demanding and least loyal customers you can attract.
• You cap your brand, nobody perceives the cheapest option as the best.
• You start a war you cannot win, someone can always go lower.

There's a quieter cost too. Cheap pricing wears out the people doing the work. When every job is priced so tight that there's no room for care, the team starts cutting corners just to survive the week. Quality drops, reviews soften, and the business ends up exactly as ordinary as its price said it was. The discount becomes a self-fulfilling prophecy.

And discounting is addictive. Run one sale and customers learn to wait for the next. Now your full price is the price only impatient people pay, and your margin depends on a calendar of promotions you can never stop running.

Why premium pricing attracts

Raise your prices with confidence and something counterintuitive happens: certain buyers want you more. Premium pricing filters out the tyre-kickers, attracts customers who value quality over cheapness, funds a genuinely better experience, and positions you as the serious choice. The goal is not to be the most expensive for its own sake, it is to charge what you are worth and then deliver like you mean it.

Here's what the extra margin actually buys you:

• Time to do the job properly. Fewer, better clients means more attention on each one, which means better results, which means referrals.
• Money to invest in the experience. Onboarding, communication, packaging, follow-up. The small touches that make people feel looked after cost money, and cheap pricing never leaves any.
• Room for marketing. When each sale is worth more, you can afford to spend more to win it. That's how premium brands outspend budget brands on attention while running leaner.
• Better customers. People who pay more tend to take the relationship more seriously. They show up, they follow advice, they value the outcome.

Some of your current customers will leave when prices go up. That's part of the plan, not a side effect. The ones who leave are usually the ones who were costing you the most in time, stress and margin.

What has to be true before you raise prices

Premium pricing isn't a trick. It's a promise, and the business has to be able to keep it. Before you change a number, check these five things:

1. Your positioning is sharp. You're clearly the best choice for a specific kind of customer with a specific problem. "Quality service at a fair price" is not positioning. We cover this properly in why positioning comes before the logo.
2. Your message matches the price. If your website, proposals and social feed look and sound like a budget operator, a premium price will feel like a mistake. The words have to carry the value before the price appears.
3. You can show proof. Results, reviews, case detail, credentials. The higher the price, the more evidence the buyer needs to feel safe.
4. The experience lives up to it. Response times, communication, presentation and follow-through all need to feel premium. The price sets the expectation, the experience has to meet it.
5. You can say the number without flinching. Buyers hear hesitation. If you apologise for your price, they'll assume you have reason to.

If two or three of those are shaky, fix them first. The price increase will land far better on a business that already feels worth it.

How to raise your prices without losing your best clients

Here's the process we'd walk a client through:

1. Work out your real costs and margins. Include your own time at a proper rate. Plenty of owners discover at least one service they're effectively running at a loss.
2. Decide who your ideal customer is. Price for the customer you want more of, not the one who complains most.
3. Repackage before you reprice. Bundle your work into clear packages with defined outcomes. A package priced on the result is much easier to lift than an hourly rate customers compare line by line.
4. Add a premium tier. Offer a top option with more speed, more access or more done-for-you. Some people will take it, and its presence makes your core offer look reasonable.
5. Apply new prices to new customers first. It's the lowest-risk test. Watch how enquiries respond before you touch existing clients.
6. Give existing clients notice and a reason. A clear date, a short explanation and, if you like, a grace period at the old rate. Loyal clients rarely leave over a well-handled increase.
7. Review every six to twelve months. Small, regular increases are easier for everyone than a big jump after five years of standing still.

A worked example: the stuck service business

Here's a hypothetical. Say you run a small bookkeeping practice in Newcastle charging $60 an hour. You're flat out, working evenings, and still not making much. Clients haggle over invoices because they can see every hour itemised, and half your week goes on three demanding clients who pay the least.

Now picture the same practice after a deliberate move up-market. You define a clear ideal client: trade businesses with a few staff who hate paperwork. You replace hourly billing with three monthly packages built around outcomes: books reconciled, BAS lodged on time, a short monthly call explaining the numbers. You add a premium tier with a same-day response guarantee. The website and proposals get rewritten to speak directly to tradies and show real examples of the headaches you remove.

New clients come in on the packages. A couple of the old hourly clients leave. The ones who stay mostly move to the middle package, because it's clearer and saves them thinking about hours. You end up with fewer clients, more predictable income and your evenings back. Nothing about the skill changed. The positioning, packaging and price did.

That's the pattern. Not a braver number on its own, a business redesigned so the braver number makes sense.

The objections you'll hear, and how to handle them

"You're more expensive than the others." Agree. Then explain why, in terms of outcomes the buyer cares about. "We are. Here's what you get that the cheaper options don't." Never argue about the number. Talk about the difference.

"Can you do a discount?" Trade, don't cut. Reduce the scope, not the rate. "We can bring the price down by removing X." That protects your value and shows the price was never arbitrary.

"I need to think about it." Usually means a question hasn't been answered. Ask what's holding them back. It's rarely really the price. It's usually risk, and risk is solved with proof and guarantees, not discounts. Our piece on building an offer they can't refuse goes deep on that.

Common premium pricing mistakes

• Raising the price and changing nothing else. The market notices the new number but not a new reason to pay it.
• Charging premium, delivering standard. The fastest route to refund requests and angry reviews.
• Hiding the price out of fear. Being vague looks like you're unsure. Confident, clear pricing reads as premium.
• Caving on the first objection. Every discount given under pressure teaches the market your prices are negotiable.
• Competing on "best" instead of "different." Trying to out-quality everyone is exhausting. As we argue in why being the best is quietly killing your business, being the obvious choice for a specific buyer is far more profitable.

Stop running the wrong race

We help brands escape the discount trap, building the positioning, message, and experience that justify premium prices and attract customers who happily pay them. If you are winning on price and losing on profit, you are running the wrong race. Let's move you up-market where the air is clearer and the margins are real.

It starts with a clear-eyed look at who you serve, what you promise and how you present it. That's the core of our consulting work: pricing, packaging and positioning built together so the number on the proposal feels like the obvious price for what the client is getting.

Frequently asked questions

How do I know if my prices are too low?

Common signs include being booked out but barely profitable, almost nobody pushing back on price, attracting clients who haggle and complain, and no money left to invest in marketing or better systems. If you win nearly every quote, you are probably underpriced. A healthy price usually means some prospects say no, because the right buyers are paying for value rather than shopping for the lowest number.

Will I lose customers if I raise my prices?

You may lose some, and that can be healthy. The customers most likely to leave are usually the most price-sensitive and demanding. If you raise prices alongside clearer positioning, stronger proof and a better experience, most good clients stay. Test new prices on new customers first, give existing clients notice with a clear reason, and watch enquiry quality as well as volume.

How much should I increase my prices by?