If your Performance Max campaign is showing a ROAS of 8, 10, or 12 — and you're wondering why your overall business performance doesn't feel like it reflects that — there's a good chance PMax is quietly eating your brand traffic and taking credit for conversions it didn't actually drive.
This is one of the most common structural problems we see in eCommerce Google Ads accounts. And it's almost always invisible unless you know exactly what to look for.
What's actually happening inside the black box
Performance Max uses every available inventory type — Search, Shopping, Display, YouTube, Gmail, Discover — and Google's automation decides how to allocate budget across all of them. That sounds fine in theory. In practice, it means PMax will aggressively pursue your brand keywords because branded search converts at extremely high rates, and the algorithm optimises for conversion volume and value, not for efficiency or incrementality.
When someone searches for "your brand name" and PMax serves them a Shopping or Search ad, PMax records that as a conversion it generated. But here's the problem: that person was going to buy anyway. They already knew who you were. PMax didn't acquire a new customer — it intercepted existing demand and took the credit.
How to tell if this is happening to you
The clearest signal is a gap between your reported PMax performance and your actual business results. If your Google Ads ROAS is strong but your revenue growth is flat, or your new customer numbers aren't moving, brand cannibalisation is worth investigating immediately.
More specific things to look for:
- PMax campaign has an unusually high ROAS compared to your non-brand Shopping or Search campaigns
- Search Impression Share data shows PMax capturing queries that match your brand name
- Your brand campaigns (if you have them) show declining impression share or click volume
- Branded keywords appear in the Insights section of your PMax campaign
- Customer acquisition costs are rising even as your ROAS looks strong
The challenge is that Google makes this difficult to diagnose. PMax reporting is intentionally opaque — search term visibility is limited, asset group data is aggregated, and there's no easy way to see exactly which queries PMax is serving. You have to piece the picture together from multiple data points.
Why Google's default setup lets this happen
By default, Performance Max has no brand exclusions applied. Google will happily spend your budget on branded queries because they convert well and make the campaign look good. This isn't a bug in the traditional sense — it's the algorithm doing what it's designed to do, which is maximise conversion value. But the incentives are not aligned with what you actually want, which is efficient new customer acquisition.
The fix requires you to take deliberate action. Google won't exclude your brand for you.
Three structural fixes that actually work
1. Apply brand exclusions to your PMax campaign
Google now allows you to add brand exclusions at the campaign level in PMax. Go into your PMax campaign settings and look for the "Brand exclusions" option. Add your brand name and any common misspellings or variations. This prevents PMax from bidding on branded search queries.
Important caveat: brand exclusions in PMax apply to Search inventory only. PMax can still target your existing customers through audience signals on Display and YouTube. This is a partial fix, but it's the most important one to apply first.
2. Run a dedicated brand campaign alongside PMax
A separate branded Search campaign, with exact and phrase match brand keywords, gives you control over your branded search presence. Set this campaign to a target CPA or manual CPC that makes sense for defensive brand spend, and make sure PMax can't compete with it directly by using the brand exclusions above.
This lets you maintain coverage of branded queries at efficient cost while PMax focuses on non-brand new customer acquisition — which is where you actually need the algorithmic help.
3. Segment and measure incrementality
The longer-term fix is building a measurement framework that separates brand from non-brand performance, and ideally measures incrementality rather than just attributed conversion value. This means looking at new customer ROAS separately from total ROAS, tracking brand search volume independently, and using hold-out experiments where possible to understand what PMax is actually generating versus intercepting.
Most accounts don't have this sophistication built in. But once you do, the decisions about budget allocation become much clearer.
What good PMax structure looks like for eCommerce
Once you've addressed the brand cannibalisation problem, Performance Max can genuinely be a powerful channel for eCommerce scaling — particularly for finding new customers at scale across Google's inventory. But it works best when it has strong signals, clear creative, well-structured asset groups, and isn't competing with your own brand campaigns for credit.
The brands that get the most from PMax are the ones that treat it as one piece of a structured account — not a black box you throw budget at and hope for the best.