Google Moved Your Search Campaigns to AI Max. Here Is What to Check This Week
AI Max widens what your campaigns match to and where they send people. That can find real demand. It can also spend your budget on queries and pages you never chose.
By Ads Ninza · Published 2 September 2026 · 8 min read
What AI Max actually changes
AI Max for Search is a set of features Google layers on top of a standard Search campaign. It became the default for new Search campaigns earlier this year, and Google has been moving existing campaigns onto it, starting with accounts that already used broad match or automatically created assets. Dynamic Search Ads are being folded into it as well.
In practice it changes three things.
Search term matching. Your keywords become a starting signal. Google uses your landing pages, your ads and its own model to match queries you did not list. On a well-built account this finds genuine demand phrased in ways nobody thought to add. On a loosely built one it finds cheap clicks that look relevant and are not.
Final URL expansion. Google can send a click to a different page on your site if it thinks that page matches the query better than the one you chose. For an e-commerce catalogue this can help. For a lead generation site it often sends people to a blog post, a careers page or the home page, none of which were built to convert.
Text customisation. Google rewrites headlines and descriptions from your site content. It is usually readable. It is not always accurate, and it does not know which claims you are allowed to make in a regulated category.
The seven checks we run after the switch
1. Search terms, daily, for two weeks. Sort by cost. Anything irrelevant goes into a negative keyword list the same day. The first fortnight is when AI Max explores most widely, and it is also when you have the most leverage to shape it.
2. Final URL expansion. Look at the landing page report. If clicks are going to pages you never built for paid traffic, switch expansion off or add URL exclusions for the pages that should never receive ads.
3. Generated text. Review every asset Google has created. Remove any that make promises you cannot keep, quote prices that are wrong, or state claims your industry rules do not allow. Healthcare, finance, legal and property advertisers should treat this as mandatory.
4. Brand and non-brand separation. Broader matching pulls branded queries into non-brand campaigns. If your brand terms are not excluded from generic campaigns, non-brand performance will look better than it is.
5. Conversion actions. Confirm the account is bidding toward a real outcome. If a newsletter sign-up or a page view is set as a primary conversion, AI Max will optimise hard toward it.
6. Location settings. Check that targeting is set to people in your locations, not people interested in them. Wider matching plus loose geography is how Indian advertisers end up paying for clicks from outside India.
7. A fair comparison. Compare cost per qualified lead, not cost per conversion, over the four weeks before and after. If you only compare platform-reported conversions, you will be grading AI Max on its own homework.
Why lead quality is the real risk
Every automated bidding and matching system does the same thing: it finds more of whatever you told it counts as success. AI Max is no different, and it has more room to do it.
If your conversion is "form submitted", AI Max will find people who submit forms. Students researching a topic, competitors checking your prices, job seekers and people who wanted a free version all submit forms. The platform sees success. Your sales team sees a longer list of calls that go nowhere.
The fix is upstream of Google. Send lead outcomes back from your CRM as offline conversions: qualified, proposal sent, won, with values. Once the platform can see which clicks became revenue, wider matching becomes an advantage, because it has a better target to aim at. Without that feedback, wider matching mostly widens waste.
This is the same work we describe on our Conversion API and signal infrastructure page, and it matters more now than it did a year ago.
When AI Max helps, and when to hold back
It tends to help accounts that have clean conversion data, a reasonable volume of conversions each month, landing pages that each cover one clear offer, and a product people search for in many different ways. E-commerce catalogues and multi-service businesses with good tracking often see new query coverage at an acceptable cost.
It tends to hurt accounts with low conversion volume, soft conversion definitions, a website where many pages are not built for enquiries, or a narrow high-value offer where one wrong query costs a lot. A B2B firm spending on a handful of expensive terms usually has more to lose than to gain.
Where the account is not ready, the practical answer is to tighten it rather than fight the setting: fix conversion definitions, exclude pages, build negative lists and restrict text customisation. Then let the wider matching run against a better signal.
How this differs by market
The mechanics are the same everywhere. The cost of getting it wrong is not.
In the USA and the UK, where clicks in legal, financial and home services categories are expensive, a week of loose matching can cost more than a month of careful management. Daily search term review during the first fortnight is non-negotiable there.
In India, clicks are cheaper but lead quality problems are sharper, because broad matching picks up a lot of informational and job-seeking traffic. Negative lists for "free", "course", "job", "salary" and "meaning" usually pay for themselves within days.
In the UAE and Canada, language adds a layer. AI Max may match Arabic or French queries to English ads, or generate text in a language your landing page does not support. Check the language of both the matched queries and the generated assets.
Frequently asked questions
What is AI Max for Search in Google Ads?
It is a set of features added to standard Search campaigns that broadens which search queries your ads can match, lets Google choose a different landing page on your site, and generates ad text from your website content. It is now the default for new Search campaigns, and existing campaigns have been moved onto it in stages.
Can I turn AI Max off?
For many accounts the upgrade itself is not optional, but the individual features can be controlled. You can switch off final URL expansion, add URL exclusions, restrict or remove generated text, and shape matching with negative keywords and brand exclusions. Controlling the parts is usually more useful than trying to avoid the whole.
Will AI Max increase my cost per lead?
It can go either way. With clean conversion data and good landing pages it often finds new demand at a similar cost. With soft conversion definitions or a site full of pages not built for enquiries, cost per qualified lead usually rises even when the platform reports more conversions. Measure qualified leads, not platform conversions.
What is final URL expansion and should I use it?
It lets Google send a click to whichever page on your site it thinks best matches the query, instead of the page you chose. For large e-commerce catalogues it can help. For lead generation sites we usually turn it off or exclude most pages, because it tends to send paid traffic to blog posts and home pages that do not convert.
Does AI Max replace Dynamic Search Ads?
Google has announced that Dynamic Search Ads are being upgraded into AI Max, with the transition running into 2027. If you rely on DSA today, plan the move deliberately: review which pages DSA was allowed to use and carry those restrictions across as URL rules.
How often should I check the search terms report?
Daily for the first two weeks after an upgrade, then two or three times a week. Sort by cost and add irrelevant queries to negative lists the same day. That fortnight is when matching is widest and when you have the most influence on what the campaign learns.
Is AI Max good for small budgets?
Small budgets have less room to absorb exploration, so waste shows up faster as a share of spend. It can still work, but the preparation matters more: strict conversion definitions, tight negative lists and one clear landing page per offer. On very small budgets we sometimes keep campaigns narrower on purpose.
Can AI Max write ads that break advertising rules?
It can generate claims that are inaccurate or not permitted in regulated categories such as healthcare, finance, legal services and property. Google policies still apply to you as the advertiser. Review every generated asset and remove anything you would not have written yourself.
Do I need an agency to manage AI Max?
Not necessarily. The checks in this article can be run by anyone who knows the account. What an experienced manager adds is judgement about which search terms are worth keeping and how to connect CRM outcomes so the system learns from revenue rather than form fills.
How do we start with Ads Ninza?
Apply through the form on this page. We start with a paid discovery call, INR 499 in India and $9 elsewhere, where a senior consultant reviews your account with you. Google Ads management starts at INR 35,000 per month in India and $500 per month outside India.