Pull the Performance Max search terms report, tag every query that contains your brand name or a misspelling of it, and see what share of PMax’s search spend and revenue sits on those terms. If brand queries carry a big slice of PMax’s revenue, its ROAS is overstating how much new demand it creates. The fix is structural: give brand searches one home (a dedicated brand Search campaign), add brand exclusions to PMax, and judge PMax on non-brand results only. Google itself now says PMax may filter out brand and navigational searches so brand campaigns can serve them, which makes the split more important, not less.
Why does Performance Max look better than it really is?
Performance Max is built to find conversions wherever they’re cheapest. Your own brand searches are the cheapest conversions in the account. These people already know you and were going to buy. If nothing stops it, PMax will happily serve on those searches and take credit for the sale.
That doesn’t break anything on its own. The problem is the decision it leads to. A PMax campaign showing 6x looks like a growth engine, so it gets more budget. But if most of that 6x came from people typing your name, the extra budget isn’t buying new customers. It’s buying sales you’d have made through a brand campaign at a lower cost, or through an organic click for free.
In September we reviewed an ecommerce account in industrial supplies spending about $7,000 a day. Brand searches drove about $59K in revenue over 30 days. Roughly 89% of that was credited to product campaigns, mostly eight PMax campaigns, not to the brand campaign. One PMax campaign looked like the best performer in the account at about $6 back per $1. Strip out the brand searches and it returned about $0.83 per $1. On paper it deserved more budget. In reality it was losing money on new demand.
How do I check how much PMax spends on brand searches?
Google added a full search terms report for Performance Max in 2025, along with a channel performance report, as part of its visibility and control updates. That gives you what you need to answer the question in about 30 minutes.
- Open the PMax search terms report for the last 60 to 90 days, with cost, conversions and conversion value.
- Export it and tag brand terms. Include your brand name, common misspellings, product line names that only you sell, “discount code,” “coupon” and “reviews” variants of your name.
- Total spend and revenue for brand vs non-brand queries. Work out brand’s share of PMax search spend and of PMax search revenue.
- Compare ROAS. Brand queries in PMax vs your brand Search campaign vs non-brand queries in PMax.
- Check the channel report. If most PMax spend is on Search placements, the brand share matters even more, because that’s where brand demand lives.
Two warnings. The search terms view only covers the search side of PMax, not YouTube, Display or Gmail. And misspellings hide well. In one audit we found brand traffic under four or five spellings of the name that nobody had tagged.
Is brand spend in PMax always a problem?
No. Sometimes the brand queries convert well and cost little. The problem is how you read the numbers.
In a September audit for a DTC hobby brand spending about $70K a month on Google, PMax’s search terms showed roughly 32% of its search spend going to brand-shaped queries: the name, misspellings and discount-code searches. Those queries returned about 10x. The brand Search campaign over the same three months returned about 8x. So PMax was doing well on brand, but it meant brand demand was split across two campaigns, and a third of PMax’s reported search performance wasn’t new category demand. We recommended a reporting split so the client stopped reading PMax’s ROAS as proof that non-brand was scaling.
That’s the core issue. Brand revenue in PMax isn’t fake. It’s just not incremental, and it shouldn’t drive budget decisions about prospecting.
What changed in 2026?
The controls have caught up a lot in the last 18 months:
- Brand exclusions. You can attach a brand list to PMax and exclude your own brand. In PMax, brand exclusions apply to Search, Shopping and YouTube search inventory, with an option to keep Shopping ads eligible on searches that mention excluded brands.
- Negative keywords. Campaign-level negatives came to all PMax campaigns in 2025 and apply to Search and Shopping inventory, per Google’s help page. Google has since raised the limit to 10,000 per campaign, matching Search.
- Search campaign priority. Google says a Search campaign with a keyword that exactly matches the query, or is spell-corrected to it, is preferred over Performance Max. If no keyword matches, the higher Ad Rank wins.
- Brand filtering. That same help page now says PMax “may filter out navigational or brand-specific search queries” so dedicated Search campaigns serve them, and tells advertisers to make sure brand campaigns have enough budget. PPC Land reported the wording on November 2, 2026. Google hasn’t announced it, defined “brand,” or said which inventory it covers.
My read on the last point: don’t rely on it. Set up your own exclusions so the outcome doesn’t depend on an undocumented filter. And if you’ve been running without a brand Search campaign because PMax “covered it,” fix that now.
How should I separate brand and non-brand?
The rule we use: every type of search gets one home.
| What the person searched | Where it should run | How to enforce it |
|---|---|---|
| Your brand name and misspellings | Brand Search (plus branded Shopping if you sell many products) | Brand exclusions in PMax, brand negatives in non-brand Search and Shopping |
| A specific product or spec | Non-brand Search, backed by Shopping | Exact and phrase keywords, product-level bids |
| A product category | Shopping and PMax, split by product line | Feed labels and asset groups by line |
| Not in market yet | A measured prospecting test (YouTube, Demand Gen) | Separate budget and its own targets |
Then run the change in this order:
- Build or check the brand Search campaign first. Exact and phrase brand keywords, enough budget that it doesn’t lose impression share to budget.
- Add brand exclusions to PMax. Use Google’s brand list for your brand, and add misspellings as negatives where the list doesn’t catch them.
- Add brand negatives to non-brand Search and Shopping. Otherwise broad match will pull brand traffic back in.
- Reset PMax targets. Without brand, PMax’s ROAS will drop. That’s expected. Set its target from non-brand reality, not the old blended number.
- Watch total brand revenue for four weeks. If brand Search and Shopping pick up the volume, you’re done. If total brand revenue drops more than spend saved, look at brand impression share before reversing anything.
Should I run a separate branded PMax campaign?
Some accounts do, and it can work. But in our experience it mostly duplicates what brand Search and branded Shopping already do, at a higher cost.
In a DTC consumer goods account we manage, the branded PMax campaign spent mostly on Search placements, the same job brand Search was doing at a better return. In August it spent about $12.8K at roughly 6x. We didn’t switch it off. We stepped it down over two months and watched whether brand Search and branded Shopping absorbed the volume. By October, branded PMax spend was down about 40% and its return had risen to nearly 9x. Across all brand campaigns, spend fell about 30% while brand revenue fell about 16%, so blended brand return went from roughly 6.7x to 8.1x. That freed budget for non-brand Search and Shopping, which is where new customers come from.
The lesson isn’t “never run branded PMax.” It’s that brand demand will find you either way, so pay as little as you can for it and put the savings into finding people who haven’t heard of you yet.
If you want a second opinion on how your brand traffic is split, our Google Ads account audit checks this first, and the free Scorecard has a Performance Max section you can run yourself. For the bigger structure question, see Performance Max vs Standard Shopping.
