Can Performance Max be profitable but not incremental?
Yes. A Performance Max campaign can report strong ROAS while generating few conversions that would not have occurred without it. This happens when PMax primarily serves users who were already going to convert through brand search, retargeting, or direct traffic. In these cases, PMax captures attribution credit without creating additional demand. The conversions are real and profitable, but the campaign is not the cause.
Why can PMax ROAS look inflated?
PMax's default configuration gives it access to branded search queries, remarketing audiences, and existing customers. These segments convert at high rates regardless of the specific ad they see. If PMax spends heavily on these high-converting, low-incremental segments, the reported ROAS reflects their natural conversion rate rather than PMax's causal impact. The algorithm optimizes for reported conversions, not for true incremental value.
How does brand demand affect PMax reporting?
Users searching for your brand by name are already in purchase intent. If PMax serves on branded queries and attributes those conversions to itself, the ROAS figure reflects the brand's demand pull rather than PMax's performance. The same conversions would likely have occurred via a direct brand Search campaign at a lower CPC. PMax inflates its reported efficiency by claiming credit for inevitable conversions.
How does remarketing affect PMax reporting?
Remarketing audiences (site visitors, cart abandoners, past customers) have higher conversion rates than cold audiences. When PMax allocates budget toward remarketing because it generates conversions efficiently, reported ROAS benefits from audience quality rather than campaign effectiveness. These users may have converted through email, organic, or direct channels if PMax had not reached them. The high ROAS comes from the audience, not the campaign.
How do you fix misleading PMax performance reporting?
First, run an incrementality test (brand exclusion holdout or geo test). Second, compare PMax ROAS against total account ROAS, not PMax in isolation. Third, review the search query categories in PMax Search Terms Insights for brand term concentration. Fourth, check audience signal overlap with existing remarketing lists. Each of these identifies different attribution inflation sources and helps you understand what PMax is actually driving.
| Signal | What It May Mean | How To Verify | Action |
|---|---|---|---|
| PMax ROAS is much higher than non-brand Search | PMax is serving brand and retargeting heavily | Check Search Terms Insights for brand term categories | Run brand exclusion test |
| Total account conversions flat after PMax launch | PMax redistributing demand, not generating new | Compare total account conversions before and after PMax | Run geo or time holdout test |
| Brand Search impression share dropped after PMax launch | PMax competing with brand Search campaign | Brand Search IS report over time | Add brand exclusions |
| Organic brand clicks declined after PMax launch | PMax capturing demand that would have come through organic | Search Console brand click trend | Evaluate brand exclusion test results |
| PMax audience composition is mostly remarketing | Budget concentrated on high-converting existing users | Audience segment report in PMax | Adjust audience signals, test cold audience asset groups |
Signals that PMax may not be incremental
- –Brand Search impression share declined after PMax launch
- –Total account conversion volume did not increase materially after PMax launch
- –PMax Search Terms Insights shows heavy brand term category presence
- –Organic branded clicks (Search Console) declined after PMax launch
- –PMax ROAS is significantly higher than brand Search ROAS (brand Search should usually be highest)
How to test incrementality
- –Add brand exclusions and run a 4–6 week holdout
- –Compare total account metrics before and after, not PMax metrics
- –Consider a geo holdout if brand exclusion test is inconclusive
- –Document results before making permanent decisions
What to do with the results
- –If total account performance holds with brand exclusions: keep exclusions, reallocate PMax budget to cold audiences
- –If total account performance drops with brand exclusions: PMax may be providing incremental brand reach; consider partial exclusions
- –If results are inconclusive: extend test or run geo holdout as second test
For detailed guidance on running tests, see our incrementality testing guide. To understand how brand and PMax interact, read about brand search cannibalization and how to audit for overlap.