Performance Max for B2B: What We’ve Learned After 18 Months

Performance Max arrived with understandable skepticism from B2B advertisers used to granular keyword-level control. Eighteen months and a meaningful number of B2B accounts later, our take has settled somewhere between the two extremes: it works, but only with structure Google doesn’t set up for you by default.

Account structure matters more than the campaign type

The biggest driver of PMax performance in B2B accounts isn’t the campaign settings — it’s whether asset groups are segmented by genuine buyer intent rather than lumped into one broad group. We structure asset groups around named use cases or industry verticals, each with its own creative and landing page, rather than a single generic group covering the whole product.

Negative signals are not optional

Without an audience signal and a consistent negative keyword list applied at the account level, PMax will happily spend budget on job-seeker and research traffic that looks similar to genuine buyer intent in the platform’s own attribution but converts at a fraction of the rate. We treat the audience signal as a starting hint for the algorithm, not a hard targeting boundary, and monitor search term reports (where available) weekly rather than assuming the black box is optimizing correctly unsupervised.

The reporting gap you have to close yourself

Google’s own PMax reporting blends channels in a way that makes true channel-level attribution difficult. We layer GA4 and CRM-level attribution on top of platform reporting for every PMax account, so budget decisions are made on pipeline data, not platform-reported conversions that may not reflect actual sales-qualified opportunities.

Where PMax still underperforms for B2B

For long, considered sales cycles with a small total addressable market, standard Search campaigns with tight keyword control still outperform PMax on cost per qualified lead. We use PMax as a complement to Search, not a replacement, for accounts fitting that profile.