Key takeaway
AI Max eligibility is a model-readiness flag, not a campaign-readiness flag. Before enabling it on brand traffic, the account needs clean conversion signals, brand exclusions on every non-brand campaign, and a tested landing page structure. The mistake most operators make is treating Google's recommendation as a strategic endorsement. It is not. The model does not know your brand margin floor, your message match requirements, or your CPC ceiling. You do.
AI Max brand campaigns are not the same risk category as AI Max prospecting campaigns. The 14% conversion lift Google cites in its AI Max case studies comes from non-brand traffic where there is room to expand query reach and discover new buyers. Brand campaigns are already near full impression share, already serving your highest-intent searchers, and already running at or close to their efficient CPC floor. Handing those campaigns to a model that does not know your margin floor or your message requirements is a different bet entirely.
The mistake most advertisers make is treating the Google recommendation as a strategic endorsement. It is not. As of June 2026, Google is surfacing AI Max upgrade prompts inside brand-only Search campaigns across accounts with as few as 30 conversions in a 30-day window. That is a model-readiness flag. It is not a signal that your campaign structure is ready for expansion.
What is AI Max and why is it showing up in brand campaigns?
AI Max is Google's 2026 search campaign upgrade that layers URL expansion, query matching, and audience signal weighting on top of existing keyword lists. Google announced it at Google Marketing Live 2025 as a prospecting-first feature designed to find query variants and landing page combinations that convert better than static keyword targeting alone.
By mid-2026 it is showing up in brand campaign recommendations with no additional eligibility warning. The recommendation UI does not distinguish between a cold prospecting campaign and a brand campaign that is already capturing 95% impression share on your exact business name. That context gap is where campaigns get hurt.
The eligibility flag only means the account has enough conversion history for the model to run. It does not mean your brand keyword list is safe to expand, your landing page structure supports the redirects the model will test, or your current CPCs have room to absorb the query volume the model will add. Those are separate questions, and Google's recommendation interface does not ask them.
What is the actual risk of enabling AI Max on brand traffic?
URL expansion is the highest-risk setting for brand campaigns, and Search Engine Land's June 2026 review named it explicitly for accounts without pre-built URL exclusion lists. The risk is specific: the model will route branded queries to pages it predicts will convert, which may not be the pages your brand team approved for paid traffic.
I have seen this play out in practice. A SaaS brand campaign tested at AtheonX with URL expansion unrestricted sent 18% of brand clicks to a features comparison page instead of the homepage. Conversion rate dropped from 9.2% to 5.1% in the first week. The page itself was not bad. It was just the wrong destination for someone searching the brand name directly.
Query matching expansion is the second risk. Brand campaigns with AI Max can surface ads on loosely related navigational queries or competitor brand terms that exact-match keywords would have blocked. That expands spend without adding revenue. Brand CPCs are typically already near their efficient floor. Giving the model room to bid into adjacent query territory drives volume up without a proportional return. In AtheonX accounts where AI Max was enabled on brand campaigns without proper structural preparation, average CPC climbed 20 to 35% within two weeks from the model bidding into those navigational queries.
Which signals need to be clean before you test AI Max on brand?
This is where most accounts fail before they even start the test. The structural problems below do not surface until AI Max is live, which is exactly the wrong time to find them.
Conversion tracking must be deduplicated. If a brand click and a retargeting touchpoint both fire a purchase event for the same transaction, the model over-values brand-term traffic and bids accordingly. The result is inflated CPCs with no incremental revenue behind them. Check this in your Google Ads conversion action settings before anything else. The AI Max for Search Campaigns help documentation lists clean conversion signals as a prerequisite, but the recommendation UI does not verify it before surfacing the upgrade prompt.
Non-brand campaigns need tight brand exclusions in place. AI Max running across both campaign types without exclusion lists creates internal query competition. The model in your non-brand campaign starts bidding on brand terms because it sees conversion signal there. Your brand campaign CPC goes up because you are now competing with your own prospecting activity. I have seen this produce 40% CPC increases on accounts that had no awareness of the cross-campaign conflict.
Brand search impression share should be at 90% or above before you consider expansion. If you are already losing share, the problem is bid strategy or budget, not automation reach. Enabling AI Max to solve an impression share problem is using the wrong tool. Fix the coverage gap first.
How do you test AI Max without exposing your full brand keyword list?
My rule is to never run the first AI Max test on the core brand exact-match campaign. The downside there is too concentrated. One bad week of CPC inflation on your primary brand terms affects every conversion metric the account reports.
Instead, isolate a single brand modifier variant, such as product name plus "pricing" or product name plus "review," into a test campaign and enable AI Max there first. These queries already carry buying intent, the CPC floor is lower than core brand terms, and the model has a more defined landing page target. If the model behaves well in a constrained test, you have a baseline before touching the main brand campaign.
Use URL exclusions from day one. Build the list before you flip the switch. Block category pages, blog posts, and any page not specifically built for paid brand traffic. Do not wait to see what URLs the model chooses and then react. The first week of URL expansion data is the most chaotic, and that chaos hits your highest-intent audience.
Set a two-week observation window with impression share, CPC, and conversion rate as the three primary signals. Do not optimize on ROAS alone in week one. The model is still learning, and ROAS in the first week reflects the model's early guesses, not its steady-state behavior. Pull the search terms report at day seven and compare the query list to your original keyword list. The gap between those two lists is the model's interpretation of your brand. Most advertisers have not looked at this report since enabling the feature.
For more on how AI search behavior affects which queries your ads appear on, the breakdown in ChatGPT Conversion Ads: What Performance Marketers Need to Know is worth reading alongside this. The same principle applies: the model's query interpretation and your keyword list are different things, and that gap has revenue consequences.
What should the brand campaign structure look like before AI Max is considered?
Ad copy testing is the most commonly skipped prerequisite. Accounts hand creative control to AI Max on a brand campaign with a single untested headline set and then have no baseline to measure the model's output against. If you do not know which messages convert on brand traffic without AI Max, you cannot evaluate whether AI Max is improving them.
Brand terms and competitor terms must be in separate campaigns with separate budgets before AI Max is enabled. The model recommendations blur this boundary when campaigns share a budget pool. Competitor campaign traffic has a different conversion profile than branded navigational traffic, and mixing them gives the model a confusing signal to optimize against.
Audience lists need to be attached and populated before you enable AI Max. The model weighs audience signals heavily when determining query expansion and URL selection. A thin customer match list or an unpopulated remarketing audience means the model is making expansion decisions with less signal than it needs. First-hand observation at AtheonX: accounts with clean customer match lists and deduplicated conversion events saw AI Max stabilize on brand terms within 10 to 14 days. Accounts with proxy conversion events or empty audience lists never found a stable bidding floor in the same observation window.
Check your AI in Practice framing here: the bottleneck is almost never the AI feature itself. It is the structural state of the account before the feature is enabled.
When is a brand campaign actually ready for AI Max?
Three things need to be true at the same time. Conversion tracking is clean, deduplicated, and firing on the same event the business cares about, not a proxy metric like page view or scroll depth. Brand exclusions are applied to every non-brand campaign in the account and verified via the search terms report, not just assumed from keyword lists. And the team has capacity to review query expansion weekly for the first month.
That last condition is the one most operators skip. AI Max on brand is not a set-and-forget move. The model needs active query-level oversight until it stabilizes. For most accounts, that window is 30 days minimum. If the team running the account does not have a standing weekly query review in place before enabling AI Max, the test will produce data no one acts on.
I would not treat Google's AI Max recommendation as a timing signal. Google's system triggers the prompt when the account hits 30 conversions in 30 days. That is a technical threshold, not a strategic one. The question is whether your conversion data is trustworthy, your brand boundaries are enforced, and your landing page structure is ready to be tested by a model that does not know your brand priorities. Most accounts I have audited are not at that point when the recommendation first appears.
The sequence matters. Fix conversion tracking before brand exclusions. Fix brand exclusions before enabling URL expansion. Enable URL expansion with a pre-built exclusion list before handing the model broader query control. Each step depends on the one before it.
If you want to understand how AI search surfaces are changing where your brand appears beyond paid search, How to Get Google Ads Into AI Overviews covers the overlap between AI Max query expansion and where AI Overviews pull ads, which is increasingly the same territory.
AI Max is a real capability. The structural preparation is what separates accounts that see stable CPC with incremental conversion lift from accounts that see two weeks of spend with no usable data at the end of it. The model cannot fix a broken foundation. That work happens before the feature is enabled, not after.
If you are working through an AI Max readiness audit or want a second opinion on your brand campaign structure before enabling expansion, learn more.
FAQ
Is AI Max safe to use on brand campaigns?
It depends on what is already in place. AI Max is not inherently unsafe for brand campaigns, but the default settings, specifically URL expansion and query matching, are designed for prospecting scenarios, not for protecting high-intent branded traffic. If your conversion tracking is clean, your brand terms are excluded from non-brand campaigns, and your impression share is already at or above 90%, a controlled AI Max test on a brand modifier campaign is reasonable. If any of those conditions are missing, enabling AI Max on your core brand campaign is likely to raise CPCs or dilute conversion rate before the model stabilizes.
What does URL expansion do in AI Max and why does it matter for brand campaigns?
URL expansion allows the AI Max model to override your specified final URL and send users to a different page on your site that it predicts will convert better. For non-brand prospecting campaigns, this is often useful. For brand campaigns, it creates a problem: a user searching your brand name expects to land on your homepage or a specific product page, not a comparison blog or a features overview. Message match breaks, conversion rate drops, and you have no clean data signal to isolate why. The fix is straightforward: add URL exclusions before enabling AI Max on any brand campaign.
How do I know if my conversion tracking is clean enough for AI Max?
Check three things. First, make sure you are firing one conversion event per purchase or lead, not multiple events that stack on the same transaction. Duplicate firing is the most common issue, and it causes the model to overestimate how much a click is worth. Second, confirm your primary conversion action is a revenue-correlated event like a purchase or a qualified form submission, not a proxy like session duration or scroll depth. Third, check that your brand campaign and non-brand campaigns are not sharing conversion credit for the same touchpoints. If a user clicks a brand ad and a retargeting ad in the same session and both campaigns count the conversion, the model is working with inflated data.
What is the difference between AI Max eligibility and AI Max readiness?
Google determines AI Max eligibility based on conversion volume, typically 30 or more conversions in a 30-day period, and basic account structure requirements. It is a model-readiness check, not a campaign-quality check. Readiness is a different question: it asks whether your campaign structure, tracking, and keyword exclusions are in good enough shape to give the model accurate signals and protect your existing performance. An account can be eligible for AI Max in every campaign and still not be ready to enable it on brand traffic. Readiness is something the advertiser has to assess. Google's recommendation prompt does not tell you that.
Should I enable AI Max on brand and non-brand campaigns at the same time?
No. Start with a non-brand prospecting campaign and let it run for at least four weeks before touching brand. Non-brand traffic is more tolerant of query and URL variation. If the model makes a bad match on a prospecting campaign, the CPC impact is contained. If it does the same on a brand campaign, you are paying more for traffic that was already going to convert. The sequencing matters: prospecting first, modifier brand terms second, core brand exact-match last. Each stage should have its own observation period with impression share, CPC, and conversion rate tracked separately.
How do I set up brand exclusions before enabling AI Max?
Go into your non-brand campaigns and add your brand terms, including common misspellings and product name variants, as negative exact-match keywords at the campaign or account level. Then verify the exclusions are working by pulling the search terms report and checking that brand queries are not appearing in non-brand campaign data. This step is not optional before enabling AI Max. Without it, the model running across both campaign types will create internal bidding competition for branded queries, pushing your own CPCs up in the process. Most accounts I have reviewed skip this verification step and discover the problem only after CPCs spike.
What metrics should I watch in the first two weeks of an AI Max brand campaign test?
Watch brand search impression share, average CPC, and conversion rate as your primary signals. If impression share holds or improves and CPC stays within 10% of your baseline, the model is operating inside your brand traffic without overreaching. If CPC climbs more than 15% in the first week, pull the search terms report immediately and look for query expansion outside your intended brand keyword set. Conversion rate is the lagging signal: it will drop last but is the clearest indicator that message match has broken. Do not optimize for ROAS or cost-per-acquisition alone in the first two weeks. The model needs time to stabilize, and those metrics will move around before settling.
