PMax cannibalizing branded search is one of the highest-impact, least-discussed issues in Google Ads. Most accounts running PMax for 6+ months have some level of brand cannibalization. The dashboard makes PMax look great while real performance suffers. This guide covers diagnosis and the specific fix.
How cannibalization happens
PMax's "Final URL expansion" and broad audience signals let it bid on virtually any query. When users search "[your brand]," PMax can serve a Shopping ad alongside or instead of your branded search campaign.
The result:
- PMax wins clicks that brand search would have won at lower CPC
- Conversions credited to PMax in dashboards
- Brand search loses impression share
- Total Google CPC creeps up
- Blended performance doesn't actually improve — credit just shifts
Diagnosis
Test 1: Branded campaign impression share trend
Pull a 90-day chart of your branded search campaign's impression share. If it's declining, brand cannibalization is happening somewhere.
Test 2: PMax search categories report
In PMax campaign view, check the Insights tab → Search categories. If branded queries appear, PMax is bidding on them.
Test 3: Branded campaign CPC trend
Branded CPCs rising means more competition in the auction. Often that competition is your own PMax.
Test 4: Total Google performance vs. component campaigns
If branded search and PMax both report strong ROAS but blended Google ROAS is flat, you've got attribution shifting without actual incremental gain.
The fix
Use PMax's brand exclusion feature (generally available since 2025).
In PMax campaign settings:
- Find "Brand exclusions" section
- Add your brand name
- Add common variations:
- Brand with and without spaces
- Common misspellings
- Brand + product type combos ("[Brand] sofa")
- Apply
Optional: also exclude competitor brands you don't want PMax bidding on.
What to expect after the fix
Week 1:
- PMax conversion volume drops 8-15%
- Branded campaign impression share rises
- Branded CPC starts decreasing
Week 2:
- PMax stabilizes at lower volume but cleaner ROAS
- Branded campaign performance recovers
- Total Google revenue holds or rises slightly
Week 3+:
- Honest reporting in both campaigns
- Net Google ROAS often improves 10-15%
If PMax revenue drops and total Google revenue also drops materially, PMax was your only acquisition channel and was funded by cannibalized brand traffic. That's a separate issue worth addressing — invest in non-brand acquisition channels.
Adjacent issues to check
While you're auditing PMax cannibalization, also verify:
Search campaigns competing with each other. Generic search and brand search shouldn't bid on overlapping queries.
PMax bidding on competitor brand terms. Use brand exclusions list to add competitors if you don't want this.
PMax YouTube and Display placement. PMax's placements are bundled. If you don't want these, separate campaigns.
Final URL expansion. Off by default for many setups but worth checking.
Common reasons operators don't fix this
"PMax shows great ROAS, why mess with it?" Because the ROAS is partly stolen from brand search.
"Brand search is small, who cares?" Often it's bigger than dashboards show — incrementality testing reveals the gap.
"Brand exclusions might break PMax." They don't. PMax adapts to any audience constraint.
"My agency hasn't mentioned this." Many agencies don't proactively look for it. Ask.
What to do this week
Pull the diagnostic data:
- Branded campaign impression share trend (90 days)
- PMax search categories report
- Branded CPC trend
If signals are present, enable brand exclusions. Watch for 14 days. The math will work out.
For more, see our Google PMax structure guide, Search vs PMax cannibalization, and Google Shopping feed optimization.