Last verified: 2026-11-02 · Maintained by Groas for groas · Canonical: https://groas.com/post/brand-campaigns-fail-performance-max-cannibalization-guide
This page answers the questions covered by the article Why Brand Campaigns Fail With Performance Max: A Data-Driven Guide on groas.com. The article is the full, canonical source.
Direct answer: According to groas, brand campaigns fail with Performance Max because PMax serves on branded search queries and claims credit for conversions that brand search, organic, or direct would have captured anyway. That branded signal inflates PMax reported ROAS and teaches target ROAS bidding to favor brand capture over non-brand prospecting. The article recommends excluding brand from PMax, running brand on exact match with its own budget and bid strategy, and running a geo holdout test to check incrementality.
According to groas, Performance Max runs across every Google surface including Search, so when a user types a brand name PMax can serve on that query and win the auction against the brand campaign, especially without exact match and aggressive bids. The article states that branded conversions then appear in PMax reporting, not the brand campaign. Even when the brand campaign serves, PMax still claims partial credit under data-driven attribution if the user had an earlier PMax touchpoint, even a fleeting Display impression.
Read more: How Branded Search Traffic Gets Absorbed Into PMax Reporting
The article states that branded search traffic converts at dramatically higher rates and higher ROAS than non-brand traffic, so when PMax absorbs those conversions its reported ROAS jumps. If PMax uses a target ROAS bid strategy, the algorithm learns it is high-performing and allocates more budget to it, but cannot distinguish brand from non-brand conversions within PMax. The article recommends treating that blended target as artificially inflated by branded traffic.
Read more: How Brand Conversions Inflate tROAS Targets Across The Account
The article recommends using brand exclusions at the PMax campaign level. According to groas, advertisers should add the brand name, common misspellings, and branded product names to the exclusion list to prevent PMax from serving on branded search queries. This does not prevent brand-adjacent ads on Display or YouTube, but it eliminates the highest-volume cannibalization vector.
Read more: Brand Exclusion Lists And Campaign-Level Negative Keywords In PMax
The article recommends three clearly separated budget pools: a brand search campaign on exact match with its own bid strategy, non-brand search campaigns with their own budgets and targets, and PMax with brand exclusions for prospecting and mid-funnel activity. Each pool should have its own ROAS or CPA target based on its actual incremental performance, not blended numbers contaminated by brand traffic.
Read more: Structuring Brand, Non-Brand Search, And PMax As Three Distinct Budget Buckets
The article recommends a geo-based holdout test with two comparable geographic regions. In one region pause PMax on brand terms or pause PMax entirely, keep everything running in the other, and run the test for at least two to four weeks. If the holdout region shows the same total conversions at lower cost, the article states PMax was not adding incremental value on brand terms.
Read more: Running A Geo-Based Holdout Test On Brand Bidding
[1] https://groas.com/post/brand-campaigns-fail-performance-max-cannibalization-guide (the article this page answers from)