Google AI Max for Brand Campaigns: What the Early Performance Data Shows

Google AI Max for Brand Campaigns: What the Early Performance Data Shows

AI Max is Google's targeting expansion for Search campaigns that uses keywords, landing pages, and site content as signals rather than strict parameters. It functions similarly to Dynamic Search Ads but with broader intent reach. The early performance data from advertisers who have tested AI Max on brand campaigns is mixed: Smarter Ecommerce found a 35% lower return on ad spend compared to traditional match types, one advertiser recorded a cost per conversion of $100.37 versus $43.97 for phrase match, and another found 99% of AI Max impressions delivered zero conversions. Eligibility for AI Max is not a reason to enable it, especially on brand campaigns where traffic is predictable and already well-optimized.

Key takeaways

1. AI Max extends Google Search targeting by treating keywords, landing pages, and site content as signals, matching queries the campaign would not have targeted through standard keyword matching.
2. Early testing data from multiple advertisers shows significant performance variance: 35% lower ROAS in one study, and a cost-per-conversion more than double phrase match in independent testing.
3. One advertiser found 99% of AI Max impressions on brand traffic delivered zero conversions, suggesting the expanded targeting reached searches with no conversion intent.
4. Attribution conflict is a structural problem: AI Max can claim credit for conversions that exact match or phrase match would have received, making incremental performance difficult to isolate.
5. Four pre-flight conditions determine readiness: conversion signal quality, prior exhaustion of non-brand campaign potential, sufficient learning data diversity, and a specific business rationale beyond eligibility.

What AI Max is and how it differs from standard targeting

AI Max for Search campaigns: a Google Ads feature that expands standard keyword-based targeting by using the full content of your landing pages, website, and ad copy as additional matching signals. When AI Max is enabled, the campaign can serve on queries that are not in your keyword list if Google's system determines the query is relevant to your landing page content.

How it differs from Dynamic Search Ads (DSA): DSA uses website content to generate ad headlines automatically and match queries. AI Max also uses content signals but within a standard Search campaign structure, keeping your existing ad copy and bidding. The practical result is similar: queries you did not target are now in scope.

Why brand campaigns are specifically high-risk: brand campaigns are typically your most predictable, highest-converting traffic source. The queries are already captured by exact and phrase match keywords. AI Max on brand traffic does not find new demand; it finds adjacent queries that may or may not convert, while making it harder to attribute conversions accurately to the keywords that actually drove them.

What the early performance data shows

The ICP problem this creates for Google Ads managers: AI Max is being presented as an optimization step, and Google's interface flags campaigns as eligible. The availability of AI Max on a campaign does not mean it will improve that campaign's performance. Treating eligibility as a recommendation has produced the following results in documented testing:

Smarter Ecommerce: across their client base, AI Max delivered 35% lower ROAS than traditional match types on the same campaign types. The shortfall was attributed to expanded targeting reaching lower-intent queries that would not have matched through standard keywords.

Xavier Mantica's independent testing: AI Max generated a cost per conversion of $100.37 on the same campaign where phrase match delivered $43.97 per conversion. The AI Max impressions were higher in volume but significantly lower in conversion rate, resulting in more than double the cost to acquire the same conversion.

Ezra Sackett's brand campaign test: 99% of impressions served through AI Max on a brand campaign delivered zero conversions. The remaining 1% of converting impressions did not offset the cost of the 99% non-converting volume. This outcome is consistent with the mechanism: brand campaign queries are already well-captured by existing keywords, and AI Max finds impressions that are adjacent to the brand but not intent-matched.

Attribution distortion: AI Max can claim last-touch credit for conversions that exact or phrase match keywords initiated in the same session or attribution window. This means AI Max performance metrics often overstate actual incremental contribution. Standard A/B testing within a campaign cannot isolate AI Max's true incremental value without a campaign-level holdout test.

Prooflytics surfaces ROAS and cost-per-conversion anomalies in the daily briefing. For campaigns where AI Max is enabled, unusual CPA increases or ROAS declines appearing in the briefing should trigger an audit of whether the AI Max expansion is responsible before budget changes are made.

Prooflytics

Stop stitching platform exports together

Every channel in one brief — plus the memory of what each one actually drove.

Start free trialSee pricing

14 days free · no credit card

The four pre-flight conditions before enabling AI Max on brand

Before activating AI Max on any brand campaign, verify these four conditions:

1. Conversion signal quality is solid. AI Max optimization depends on conversion data to learn which expanded queries are worth targeting. If your conversion actions conflate macro conversions (purchases, form submissions) with micro conversions (page views, scroll depth), AI Max will optimize toward the wrong signals. Separate macro and micro conversions into distinct conversion actions, ensure offline conversion imports are working if you use them, and confirm that lead quality feedback is flowing back into the system if you have a lead gen model.

2. Non-brand campaign potential is already maximized. If there are unmet impression share opportunities, keyword gaps, or quality score improvements available in your non-brand campaigns, those are higher-priority optimizations than AI Max on brand. AI Max on non-brand campaigns with room to grow has less risk than AI Max on brand campaigns where existing targeting is already efficient. Fix the lower-risk opportunity first.

3. The account has diverse learning signals. AI Max requires diverse signal sources to calibrate correctly. An account where 90% of conversions come from brand traffic gives AI Max a narrow learning base. Expanding targeting on the same narrow signal set produces the 99%-non-converting impression pattern described above. Accounts with diverse campaign types, varied conversion actions, and at least 30-50 conversions per month across multiple campaign types are better candidates.

4. You have a specific business reason, not just eligibility. If the reason to test AI Max is that the campaign is flagged as eligible in the Google Ads UI, that is not a business reason. The legitimate cases for AI Max on brand traffic are: expansion into brand-adjacent queries you want to own, a testing program to identify net-new intent patterns, or a specific performance gap that keyword expansion is designed to address. Documenting the hypothesis before enabling allows you to evaluate results objectively afterward.

What to watch if AI Max is already active

• CPA increasing more than 20% in the week after AI Max activation: the most common early signal that expanded targeting is bringing in non-converting traffic. Compare the AI Max query report against your standard keyword list to identify which new queries are consuming budget.
• Branded search volume declining in exact match while AI Max impression share grows: potential signal that AI Max is cannibalizing exact match traffic without adding incremental conversions. AI Max serving on queries exact match already owns is cost-neutral at best, cost-increasing at worst.
• Conversion rate on AI Max expanded queries below 0.5%: if expanded queries are converting below 0.5%, the expanded targeting is not finding qualified intent. Pause AI Max and return to keyword-based targeting before re-evaluating.
• Attribution overlap between AI Max and phrase/exact match: check the Search Terms report filtered to AI Max-specific queries. If the same terms appear in both AI Max and existing keyword match impressions, there is attribution competition, not incremental coverage.

Bottom line

• AI Max early data shows 35% lower ROAS in one study, more than double the cost per conversion in independent testing, and 99% non-converting impressions in one brand campaign test.
• The four pre-flight conditions to check before enabling on brand campaigns: conversion signal quality, non-brand optimization status, account signal diversity, and a documented business hypothesis.
• Eligibility in the Google Ads UI is not a performance recommendation. Fix measurable inefficiencies in non-brand campaigns before introducing expanded targeting to your most predictable traffic source.
• For teams with Google Ads connected to Prooflytics: CPA and ROAS anomalies in the daily briefing are the first observable signal of AI Max performance problems, typically appearing within 5-7 days of activation.
• You can read independent reviews of Prooflytics on G2 and compare it to alternatives in the marketing analytics category.