What you'll learn in this article
- What AI Max for Search is and why it is a layer on top of Search campaigns, not a new campaign type
- How the new format works: AI Max search term matching and AI Max asset optimization in practice
- How AI Max final URL expansion changes where your traffic lands
- When to test AI Max for Search: the account conditions that make a test worthwhile, from operational experience
- Where official Google documentation describes AI Max and where account experience adds concrete observations
- How to read the results once AI Max for Search is live, and what can go wrong
AI Max for Search is the most significant change to how a standard Search campaign behaves in years, and the most misunderstood. It is tempting to read the name as a new campaign type sitting next to Performance Max, but that framing is wrong. AI Max for Search is an optimization layer you switch on inside an existing Search campaign, and once it is on it quietly changes which queries you match, what your ads say, and where users land. The format does not replace your keywords, it sits around them.
This matters because the decision to turn it on is not the same kind of decision as launching a new campaign. You are changing the rules of a campaign that already has history, conversion data, and a learned bidding model. In this article I walk through how AI Max for Search actually works, what each of its features does to a live campaign, and the account conditions under which testing it is worth the risk, based on what I have seen running real Search accounts rather than on the launch slides.
What is AI Max for Search
AI Max for Search is a suite of targeting and creative features that you enable at the campaign level inside a standard Search campaign. The two pillars are AI Max search term matching, which broadens the set of queries your ads are eligible for, and AI Max asset optimization, which generates and adapts your headlines, descriptions, and landing pages in real time. Around these sit final URL expansion and a set of brand and URL controls. Crucially, it is one toggle that activates a bundle, not a single isolated setting.
Why it is a layer, not a campaign type
The distinction that trips up most advertisers is that AI Max for Search does not create a separate campaign. You keep your keywords, your bidding strategy, your budget, and your conversion history. What changes is how the system interprets that setup. Google's own documentation frames AI Max as an evolution of Dynamic Search Ads and of Performance Max's final URL expansion, folded back into a regular Search campaign. The official overview of the feature is here: About AI Max for Search campaigns.
From operational experience: treating AI Max for Search as a campaign type leads to the wrong test design. Because it inherits the existing campaign's history, the right comparison is not "new campaign vs old campaign" but "same campaign before and after the layer is on." That is also why the cleanest way to evaluate it is an experiment that splits traffic on the same campaign rather than a brand new build.
What the new format actually adds
On top of your manual keywords, AI Max for Search adds query coverage you did not explicitly buy, creative variations you did not write line by line, and landing page routing you did not hard-code. The trade is reach and relevance in exchange for control. The more your previous setup leaned on tightly scoped exact-match keywords and pinned assets, the more AI Max search term matching and AI Max asset optimization will change behavior, for better or worse depending on the account.
How the new format works
AI Max for Search operates through a small number of features that interact. Understanding each one separately is what lets you predict how a given campaign will respond when you flip the toggle. Google's technical breakdown of these mechanics is documented here: How AI Max for Search campaigns works.
1. AI Max search term matching
The first pillar is AI Max search term matching. Instead of being limited to the queries your keywords can reach through broad, phrase, and exact match, the system uses Google AI to find additional relevant queries based on your landing pages, ads, and keywords. Turning on AI Max turns on search term matching for the campaign, and it can be toggled at the ad group level. In practice this means your impression volume on long-tail and novel queries can rise quickly, which is exactly where reach-constrained campaigns gain.
From experience: the first thing that moves after enabling AI Max search term matching is the search terms report, which fills with queries you never added. Negative keywords are still respected, so the lever you keep is exclusion, not inclusion. On accounts with thin negative lists, the early days can surface irrelevant traffic, which is why a ramp-up period before judging the test matters.
2. AI Max asset optimization
The second pillar is AI Max asset optimization, a panel of campaign-level settings that tailors creative messaging at scale. It includes text customization, formerly known as automatically created assets, which generates headlines and descriptions from your domain, landing pages, existing ads, and keywords. The intent is that the ad copy adapts to the specific query rather than staying fixed, which lifts relevance on the broader set of terms that search term matching now reaches.
From experience: AI Max asset optimization interacts with pinning in a way that surprises people. If you enable text customization together with final URL expansion, pinning of responsive search ad assets is not respected, and pinning is also not respected when URL inclusions are provided. For brand-sensitive accounts that rely on pinned legal or brand assets, this is the single most important behavior to test before committing.
3. Final URL expansion and landing page routing
The third feature is final URL expansion, which lets the system send a user to a more relevant page on your domain than the one you set, including dynamic landing pages matched to the query. This raises landing page relevance but introduces a routing risk: combined with existing tracking templates, the dynamic URLs can produce broken expanded pages with a 404 error if the tracking templates are not configured compatibly. The new format introduces landing page automation that substitutes advertiser-provided URLs with dynamic ones.
4. The auction interaction with Performance Max
Because AI Max for Search broadens matching, it can compete with Performance Max for the same query. The rule is the same logic that governs broad match and Performance Max: if a query is eligible in both, the candidate with the highest Ad Rank serves. That means turning on AI Max in a Search campaign that overlaps a Performance Max campaign does not create duplicate serving, but it can shift which campaign wins a given query, and that shift is worth watching during a test.
When to test AI Max for Search
The decision to test AI Max for Search is an account-fit question, not a yes-or-no one. The features that make it powerful are the same ones that make it risky on the wrong campaign. Before enabling anything, Google's setup guide walks through the exact toggles and the conversion behavior of Dynamic Search Ad groups: Set up AI Max in Google Ads.
Good candidates for a test
The campaigns that gain most from AI Max for Search are reach-constrained: they have strong conversion performance but are capped by keyword coverage, leaving budget unspent or impression share lost to limited query reach. They also need stable, trustworthy conversion tracking and enough conversion volume for the system to learn from, because the value of the format comes from Google AI optimizing toward conversions on the expanded query set. Clean negative keyword lists are the third prerequisite, since exclusion is your main steering lever.
From experience: the clearest wins I have seen came from campaigns that were already converting well but were impression-share-limited on a narrow keyword set. Switching on AI Max search term matching let them capture adjacent intent they were structurally blind to before, and asset optimization kept relevance acceptable on that new traffic. The lift was real precisely because the constraint being relaxed was reach, not creative quality.
When to hold off
Hold off on AI Max for Search when the campaign is brand, legal, or compliance sensitive and depends on pinned assets or strict match control, because text customization plus final URL expansion can override pinning. Hold off when conversion tracking is shaky, since the entire value proposition rests on the system optimizing toward reliable signals. And hold off on campaigns with very thin budgets or sparse conversions, where there is not enough volume for the expanded matching to learn anything useful before the budget is spent on exploratory traffic.
How to run the test cleanly
The cleanest way to evaluate AI Max for Search is a campaign experiment that splits traffic, so the comparison is the same campaign with and without the layer. Give it a ramp-up period of at least a couple of weeks before judging, because search term matching and asset optimization both need time to learn before the numbers stabilize. Watch the search terms report for relevance, watch CPA or ROAS against your control, and watch the landing pages report for any 404 routing from final URL expansion.
Operational implications: what changes once AI Max is live
Enabling AI Max for Search is not a settings tweak you forget about. It changes the daily texture of managing the campaign: what you monitor, what controls still work, and where the failure modes are. These are the practical shifts to expect.
Your steering moves from inclusion to exclusion
Once AI Max search term matching is on, you no longer fully decide which queries you enter by choosing keywords. The system decides, and your main lever becomes negative keywords. This inverts the daily workflow: the search terms report stops being a passive record and becomes the primary place you actively prune. Accounts that treat negatives as a one-time setup will drift; accounts that review them weekly keep AI Max relevant.
Pinning and brand control become conditional
The most important operational implication of AI Max asset optimization is that creative control is now conditional. With text customization and final URL expansion both on, pinning is not respected, and it is also dropped when URL inclusions are provided. If pinning is non-negotiable for legal or brand reasons, the practical answer is to avoid final URL expansion and URL inclusions on that campaign, or to keep AI Max off there entirely.
Landing page routing becomes something to audit
With final URL expansion live, the page a user lands on is no longer guaranteed to be the one you set. That makes the landing pages report part of your routine audit, not an afterthought. The specific failure mode to watch is the 404 that can appear when dynamic expanded URLs collide with tracking templates that were not built for them. Catching that early is the difference between a clean test and a quietly leaking one.