Google Ads · 15 min

PMax Brand Cannibalisation Teardown

How an efficient-looking Performance Max campaign can capture branded demand and overstate incremental acquisition.

Direct verdict

Do not pause Performance Max because brand traffic exists. Separate the demand, apply the appropriate control, and judge the change on total-account non-brand and new-customer outcomes.

Use this resource when

PMax with unusually high ROASAccounts where brand Search volume fellTeams unable to separate new and existing demand
01

The symptom

PMax reports the best ROAS in the account while brand Search volume or conversions decline. Asset-group reporting looks healthy, but search categories and customer mix suggest the campaign is harvesting existing demand.

02

The evidence stack

Pull PMax search categories or query insight, brand Search trends, new-customer share, returning-user conversions and total-account conversion volume. Add Merchant Center brand data for commerce accounts. No single report proves cannibalisation.

  • Brand query exposure
  • Brand Search impression and conversion trend
  • New versus existing customer share
  • Total account non-brand conversions
  • Change in direct and organic branded demand
03

The controlled change

Create or confirm the correct brand list, apply brand exclusion to the relevant PMax campaign, annotate the change date and keep other material controls stable. Google supports brand exclusions for Performance Max; use negative keywords for essential irrelevant terms, not as a substitute for brand governance.

04

The measurement window

Read at least one full conversion-lag cycle and compare the same weekday mix. Watch total-account conversions and value, non-brand acquisition, new-customer share and marginal CPA—not PMax ROAS alone.

05

The decision

Keep the exclusion if total-account acquisition quality is stable or improves and non-brand contribution becomes clearer. Revisit the design if total-account qualified volume falls materially without a compensating quality improvement.

06

What I would not change

I would not pause PMax, restructure asset groups and change bidding in the same test. Multiple simultaneous edits destroy the evidence needed to decide whether brand control helped.

Historical Google Ads account overview with clicks, conversions and cost
Evidence, not decoration.Historical Google Ads evidence. A top-line account trend is the control chart for a PMax experiment: the decision must be based on total-account movement, not one campaign's reported ROAS.

Practical questions

Questions that change the decision

Does all branded traffic in PMax equal cannibalisation?

No. The question is whether the campaign receives credit for demand that would have converted elsewhere and whether controlling it improves incremental acquisition.

Should competitor brands also be excluded?

Only when they are irrelevant, create policy risk or fail the economics. Competitor intent can be valuable in some categories.

Can I use negative keywords instead?

Google recommends brand exclusions for brand searches. Use negative keywords for essential irrelevant terms or brand-safety needs.

Sources and platform documentation

Platform features change. These links point to the primary documentation used for the implementation details above.

  1. Google Ads: brand settings for Search and Performance Max
  2. Google Ads: apply brand exclusions
  3. Google Ads: negative keyword guidance

Free 48-hour audit

Turn this resource into an account-level decision.

Share your account context and bottleneck. I’ll identify the three highest-impact opportunities—without a sales deck.

Request your audit

Free operating template

Stop reviewing paid ads with screenshots and green arrows.

Use the same weekly review structure I use to connect spend with qualified leads, opportunities, pipeline and decisions.

  • Commercial scorecard
  • Creative test log
  • Decision ownership
Get a free audit