Your business does not have an AI problem. It has a calendar problem, and the numbers behind it are worse than most owners realise.
A new AI platform can easily take weeks to onboard. And Typeface's 2026 Signal Report found roughly four in ten senior marketing leaders now consider three to four weeks an acceptable timeline to launch a single campaign.
Sit with that. Weeks to learn the tool. Weeks more to ship the thing the tool was supposed to accelerate.
Buy four tools a year at a month of setup each and you have spent a third of your working year in setup, watching demos, importing data and reading documentation, while the competitor down the road ran twelve campaigns with worse software.
So here is the short answer. The value of a marketing tool is not what it can do. It is how fast it produces a real result inside your business. Measure time to value, not features, and most of the software on your wishlist either earns its place in a fortnight or never deserved it.
Why onboarding takes so long
It is rarely the software. It is four things that have nothing to do with the interface.
The data is not ready. Every tool needs clean inputs. Customer records in one format, conversion events defined properly, product data structured. Most businesses discover their data is a mess only when something tries to read it, so the first weeks of many onboardings are actually a data cleanup project wearing a different name.
Nobody owns it. The tool gets bought by one person, championed by another and used by nobody, because implementing it was never anyone's actual job with actual hours attached.
The goal was vague. "We should be using AI" is not a goal. Without a specific outcome, there is no way to know when onboarding is finished, so it never is.
It replaced nothing. The new tool was added on top of the existing process rather than replacing part of it, so now there are two ways of doing the thing and the team keeps using the old one.
Every tool you add without removing something is a tax on attention. Buy four and you have not multiplied your capability. You have quartered your focus.
What the trap actually costs you
The subscription fee is the smallest line on the bill. That is why the trap is so easy to walk into. The sticker price looks trivial next to a salary, so nobody runs the real numbers.
Here is what actually gets spent.
- Setup hours. Every hour your best people spend in demos, imports and settings screens is an hour they are not writing offers, briefing creative or talking to customers.
- Delayed campaigns. Work that could have shipped with the old process sits waiting for the new one to be ready. The market does not pause while you configure.
- Switching friction. Every login, every new dashboard and every tool with its own logic adds a little cognitive load to every task. Across a team, that load is real.
- Abandoned investment. The tool that never quite got adopted still took money, hours and goodwill. Worse, it makes the team cynical about the next one, even if the next one is genuinely good.
None of that shows up on the invoice. All of it shows up in how much work actually leaves the building.
The two week rule
Here is the constraint worth adopting, because it forces every other decision to be made properly.
Any new tool must produce one real, measurable result within two weeks of purchase. Not a completed setup. Not a training session. A result. A campaign live, a report replacing a manual one, an hour a week genuinely saved and verifiable.
If it cannot, either the tool is wrong for you or the problem is not ready to be solved with software.
This rule kills the most expensive failure mode in marketing technology, which is not buying the wrong tool. It is buying the right tool at a moment when the business cannot absorb it.
Two weeks is deliberately tight. It stops you buying the enterprise platform when you needed a single feature. It stops the vendor's onboarding programme becoming your project plan. And it forces the most honest question in software buying: what is the smallest useful thing this tool can do for us by the end of next week?
The pre purchase checklist
- Name the specific job. Not "improve our marketing." One job, done by a named person, taking a known number of hours per week.
- Name what it replaces. If nothing comes off the list, the tool is an addition, not an improvement.
- Check the data it needs. Before you buy. If your customer records are in three spreadsheets and a shoebox, fix that first or the onboarding becomes a data project.
- Assign an owner with hours. Four hours a week for the first month, blocked in the calendar. Unowned tools die quietly and expensively.
- Define the two week result. In writing, before the credit card comes out.
A useful filter before any of this: is the job worth automating at all? Some work should stay in human hands no matter how good the software gets. We break down where that line sits in what to hand to AI and what never to. If the job fails that test, no onboarding plan will save the purchase.
A worked example: the two week rule in practice
Say you run a Brisbane physio clinic with three staff. Your practice manager spends roughly five hours a week pulling booking numbers, ad spend and enquiry sources into a spreadsheet so you can see what is working. A vendor pitches you an AI reporting platform that promises to automate all of it and add forecasting, sentiment analysis and a dozen other features.
The old way, you sign up, book the onboarding calls, connect every data source and wait until the whole thing is "properly set up" before relying on it. Five weeks later it is still half configured, the practice manager is running the old spreadsheet in parallel, and nobody is sure which numbers to trust.
The two week way looks different.
- Day one. Write the job down. "Replace the Monday spreadsheet. Bookings, spend and enquiry source, one screen, updated automatically." Name the practice manager as owner, with four hours blocked this week and next.
- Days two to four. Connect only the three sources that spreadsheet uses. Ignore forecasting and every other feature for now.
- Days five to ten. Run the new report next to the old one. Check the numbers match. Fix whatever does not.
- Day eleven. Stop doing the spreadsheet. The old process comes off the list.
- Day fourteen. Judge it. Did it save the hours? Are the numbers trusted? If yes, keep it and pick the next feature. If no, cancel inside the trial and you have lost two weeks, not two months.
Notice what happened. The tool did not get simpler. The scope did. That is almost always where the time goes.
The three to four week campaign launch
The second statistic is the more damaging one, because it applies to every campaign rather than every purchase.
Roughly four in ten marketing leaders in that survey now accept three to four weeks as a normal time to launch a single campaign. In an environment where creative can wear out within weeks, that is fatal arithmetic. Your campaign is stale before it ships. If that sounds dramatic, read why a winning ad now dies in about fourteen days.
The delay is almost never production. It is the queue. Waiting for approval. Waiting for the photo. Waiting for the copy review. Waiting for someone to remember to send the file.
Three fixes, in order of impact.
Pre approve the boundaries, not the assets. Agree in advance what claims can be made, what the offer is and what the brand rules are. Then anything inside those boundaries ships without another meeting. Approval per asset is what turns a two day job into a three week one.
Shoot in batches, ahead of need. The most common cause of delay is waiting for footage that does not exist yet. A quarterly shoot day removes that dependency entirely. The one shoot, ninety days of content method shows how to plan that day so it feeds a whole quarter.
Ship at 80%. In our experience the last 20% of polish costs more calendar time than the first 80% and, when creative wears out fast, adds little to performance. A good ad live on Tuesday beats a great ad live in a fortnight.
How to measure time to value
You cannot fix what you do not measure, and most businesses have never timed their own process. Start tracking three numbers. None of them needs new software. A shared spreadsheet is plenty.
- Days from purchase to first result for every tool you buy. Write the purchase date and the date of the first genuine result. Anything past fourteen days gets a hard conversation.
- Days from brief to live for every campaign. Log when the brief was agreed and when the ad went live. Then log where it waited. You will see the same bottleneck again and again.
- Hours saved per week, verified. Not estimated by the vendor. Measured by the person who used to do the work, before and after.
After a quarter of tracking, you will know which tools pay their way and which step in your process eats the most calendar. That is worth more than any feature comparison.
What to do with the time you get back
If you cut onboarding from a month to a fortnight and campaign launch from four weeks to one, you have recovered something like six to eight weeks of working time a year.
Do not fill it with more tools. Fill it with the two things that actually compound. More genuinely distinct creative in market, and more conversations with customers.
Both are unfashionable, neither has a dashboard, and together they outperform every piece of software you were considering.
The audit worth doing today
List every marketing tool you pay for. Next to each, write the specific job it does and the last date somebody used it.
Many businesses find two or three subscriptions nobody has opened in months, and one tool that was bought to solve a problem the business no longer has.
Cancel those. Take the money and the attention, and point both at the one tool that is genuinely working.
Then sort out the process around it. That is usually where an outside view helps most, because the people inside the business are too close to the queue to see it. Our consulting work starts with exactly this kind of audit: what you pay for, what actually ships, and where the weeks disappear.
Focus is the scarce resource now. Software is not.
Frequently asked questions
How long should it take to onboard a new AI marketing tool?
A new marketing tool should produce one real, measurable result within about two weeks. Full adoption can take longer, but if nothing useful has shipped by then, the scope is too big, the data is not ready or nobody owns it. Narrow the first job to one task a named person already does, and judge the tool on that alone.
How do I know if my business is ready for an AI tool?
You are ready when you can name the exact job it will do, the person who owns it, the process it replaces and the data it needs, and that data is already clean. If your customer records live in several spreadsheets, or nobody has hours to implement it, fix that first. Otherwise the onboarding turns into a data cleanup project.
Why do marketing campaigns take so long to launch?
Mostly because of waiting, not making. Assets sit in approval queues, copy waits for review and campaigns wait for footage that has not been shot. Pre approving the boundaries, such as claims, offer and brand rules, lets work inside them ship without another meeting. Batching shoots ahead of need removes the other big delay.
How many marketing tools does a small business actually need?
Fewer than most have. A small business usually needs something to run ads, something to send email, a website it can edit and a way to see where customers came from. Beyond that, every tool should replace a specific task. If a subscription has not been opened in months, cancel it and put the attention back into the tools that are working.