Check your Google Ads account. There is a decent chance a campaign you built three years ago was converted into something else in September while you were busy running your business.
Google is auto upgrading older Search setups to AI Max. Campaigns using automatically created assets or the campaign-level broad match setting went first, from September 2026. Dynamic Search Ads follow in February 2027. Over at Microsoft, AI Max is now switched on by default for every new Search campaign, though you can turn it off.
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 your budget, your negatives and your message before you let it run the mechanics. Handled that way it can beat manual management. Handled carelessly it buys you cheap leads that never close.
You can switch individual features off. You cannot switch off the direction. Both platforms have decided that keyword level control belongs to a machine now, which leaves you one real decision: which levers you hold onto while the rest goes automatic.
What AI Max does to your account
Strip away the branding and AI Max is three features in one bundle.
Search term matching broadens your reach well past your keywords, learning from the keywords, ads and landing pages you already have. Text customisation writes headlines and descriptions from your existing ads and your landing page copy. Final URL expansion picks which page on your site each searcher lands on, even if you specified a different one.
The September upgrade does not switch on all three. Google's announcement says campaigns on the broad match setting get search term matching only. Campaigns using automatically created assets get search term matching plus text customisation. Final URL expansion stays off for both until you turn it on yourself. The full set arrives by default only for Dynamic Search Ads campaigns, and that is the February 2027 wave.
For some accounts that is a gift. If you sell one thing to a broad market and every enquiry is worth roughly the same, handing over the mechanical work frees up your week. The same Google announcement claims an average of 7% more conversions at a similar cost per conversion. Remember who is quoting it.
For others it is a slow disaster. 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 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, drift is expensive.
- The landing pages are vague. AI Max learns from your pages, so a page listing every service you have ever offered invites every search you never wanted.
Recognise your account in two or more of those? Then treat the levers below as compulsory.
Automation optimises toward the target you set. If your target is "a form submission" it will find you an endless supply of people who fill in forms and never buy.
The four levers that must stay yours
One. The conversion definition. Most accounts lose control right here. Tell Google Ads that a submitted contact form is the goal and it will deliver contact forms, cheerfully, at a lower cost per lead than you have ever seen. Your sales team then spends three weeks talking to people who wanted a free quote for something you do not sell.
Feed it qualified events instead, such as a booked appointment, an accepted quote or a sale with its dollar value attached. Google Ads has offline conversion imports for exactly this job, and in my view far too few small accounts use them. Once the import is running, the automation is hunting for the customer you want.
Two. Budget and the pace of change. Automated bidding learns by spending your money. Decide in advance how much tuition you are willing to pay, say four weeks at your current daily budget. Change one thing at a time. Fiddle every second day and you will never find out what worked.
Three. Negative terms and brand safety. Broad matching pulls in queries you would never have chosen. Some are gold. Others come from people hunting 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. I rate it the most valuable hour you can spend in the account.
Four. The message and the landing experience. Text customisation builds ads out of the copy you have already written. It cannot invent your offer. Hand it a headline like "Fixed price quote in 48 hours" and it has something worth assembling. Hand it "Quality service, great prices" and you get the same mush, faster.
The migration checklist
Work through this before you judge the results.
- Fix your conversion tracking first. Confirm what counts as a conversion and remove the duplicates. Page views and phone clicks are not sales, so stop counting them as if they were.
- Import your real outcomes. Push qualified leads and closed revenue back into the platform so the bidding learns from money, not form fills.
- Check which features are on. Open the campaign and ad group settings and note which of the three AI Max features each campaign has. Leave final URL expansion off until your pages are ready for it.
- Clean your landing pages. A vague page produces vague matching. Make each page plain about what it sells and to whom.
- Set up a search terms review. 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 campaign was built years ago around keywords like "kitchen renovation Canberra" and "custom kitchen quote", with the campaign-level broad match setting switched on. A typical job is worth tens of thousands of dollars. Your conversion action is "contact form submitted".
In September the campaign is upgraded and search term matching comes on. Within a few weeks the dashboard looks fantastic, with conversions up and cost per conversion 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. First, the primary conversion becomes "site measure booked", and signed contracts are imported from your CRM with their value. Next come negatives for jobs, DIY, flat-pack, hinges and repairs. The landing page is rewritten to say plainly that you do full kitchen renovations, in the Canberra region, from a stated starting price. Then you hold the budget steady for a month while the bidding relearns.
Cost per lead goes up, and for a week that feels bad. But the people enquiring are now planning a real renovation, and cost per signed job heads the right way. That number pays your bills. Same automation, different result, because you changed what it was hunting for.
The branded search question
Broad automated matching loves branded searches, because they convert brilliantly and make the whole campaign look like a triumph.
Most of those people were looking for you by name and would have found you anyway. If you are not separating branded from non branded performance, the automation can spend a growing share of your budget on traffic you already owned, and the dashboard will glow.
Split it out and judge the two separately. AI Max comes with brand inclusion and exclusion settings to help you do it.
How to read the results without fooling yourself
Automated campaigns come with automated reporting, written by the same company that wants more of your budget.
- Judge on revenue. 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 as well as the totals. Totals can look healthy while the mix shifts toward cheaper, worse queries.
- Give it time, then be decisive. A few weeks is too short 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.
What to do this week
- Log in and check every Search campaign. Find out which ones have been upgraded, which are scheduled, and which AI Max features are on in each.
- Open your conversion settings. Write down exactly what each campaign is optimising toward. If it is form fills or page views, flag it.
- Pull last month's search terms report. Highlight every query you would never have chosen. Add the obvious ones as negatives today.
- Split branded from non branded. Look at the non branded numbers on their own. That is your real performance.
- Book the diary slot. One hour every week for eight weeks to go through search terms and results.
- 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 no scam, and resisting it entirely is no strategy. On clean signal, with a decent offer and strong landing pages, it can beat plenty of manual account management. The machine 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 and keep the judgement. 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. 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 bundle of three automation features Google applies to Search campaigns. Search term matching broadens reach beyond your chosen keywords, text customisation generates headlines and descriptions from your ads and landing pages, and final URL expansion chooses which of your pages each searcher lands on. The September 2026 auto upgrade turns on search term matching, plus text customisation for campaigns that used automatically created assets. Final URL expansion stays off unless you enable it.
Can I opt out of AI Max in Google Ads?
You can turn the individual AI Max features off in campaign and ad group settings, but the direction is clear: Google and Microsoft are both moving search advertising toward automation. 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 a qualified event such as a booked appointment or a sale with a dollar 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.