Performance Max Brand Cannibalization: Is PMax Taking Credit for Sales You Already Owned?

Your Performance Max campaign says it's your best performer. A study of 503 Google Ads accounts says you should look closer. Optmyzr found that 91.45% of those accounts had keyword overlap between their Search campaigns and PMax, and more than half of all Search campaigns in the data set were competing with PMax for the same terms. Some of that overlap is harmless. A lot of it is Performance Max quietly taking credit for sales you would have won anyway, then reporting a return on ad spend that looks better than the reality.

We see this constantly. Across our client portfolio, the single most common reason a brand's blended Google Ads ROAS looks strong while total revenue stays flat is brand cannibalization inside Performance Max. The campaign isn't lying on purpose. It's doing exactly what Google built it to do: find the highest-probability conversions and claim them. The catch is that the highest-probability conversion in any account is almost always someone typing your brand name into the search bar, and those people were going to buy from you regardless.

How Performance Max captures traffic you already owned

Performance Max runs across every Google surface at once: Search, Shopping, YouTube, Display, Discover, Gmail, and Maps. When someone searches your brand name, that query is eligible for both your Search campaigns and PMax. For a long time, PMax won those auctions more often than not, because Google's system knows brand searchers convert at a high rate and prices the bid to win them.

Two mechanisms drive most of the damage. The first is brand search itself. Someone already knows you, searches for you, and PMax serves the ad and books the sale as PMax performance. The second is Final URL Expansion, which lets PMax match queries you never targeted to whatever landing page it decides fits. Left on, it behaves like a dynamic search campaign bolted onto PMax, pulling in both brand and long-tail non-brand traffic that your other campaigns were already covering.

The result is a blended number that flatters the channel. Adalysis, in a separate study of roughly 3,300 campaigns, found that when both PMax and Search were eligible for the same terms, the Search campaigns usually had the higher conversion rate. Put plainly: the traffic PMax absorbs often performed better when a dedicated Search campaign handled it. You're not getting more sales. You're moving credit from one line item to another and paying a premium for it. One benchmark dataset put standard Shopping's cost per acquisition roughly $5 below PMax's, $38.87 versus $43.91, on the same kind of traffic.

Why brand cannibalization costs more in 2026 than it used to

This has always been a quiet tax. It's a louder one now, because the rest of the account got more expensive. Across 2025, ROAS fell in 13 of the 14 ecommerce categories WordStream tracks, with the average dropping around 10% to roughly 3.68. Average Search CPC climbed about 12% year over year to $2.96 in early 2026. When clicks cost more and returns are thinner, every dollar PMax spends re-buying a customer you already had is a dollar that isn't finding a new one.

At the scale we operate, managing more than $7M in ad spend across 100+ brands, a few points of wasted brand spend compound into real money fast. Take a brand doing $5M a year on Google with 15% of its PMax budget quietly flowing to brand terms it would have captured through organic listings or a cheap brand Search campaign. That's a meaningful sum lit on fire every month, and the fix costs nothing but attention.

How to audit brand cannibalization in your account

The good news is that PMax stopped being a black box in 2025. Google rolled out a set of reports that finally let you see what the campaign is actually doing, and they're the foundation of any honest audit.

Start with the channel performance report, which finished rolling out across accounts in late 2025. It shows how your PMax budget splits across Search, Shopping, YouTube, Display, and the rest. If a large share is going to Search, that's your first signal that brand and text queries are in play rather than Shopping. Next, open the Search Terms report, which now covers Search and Shopping placements inside PMax. Look for your own brand name and its close variants. If they're driving a big chunk of conversions, PMax is feeding on demand you created somewhere else.

The test we run is simple: pull brand terms out of the blended number and look at what's left. If PMax ROAS drops sharply once brand is removed, then the campaign's real job, finding new customers, is weaker than the dashboard suggested. That gap is the number that actually matters, and it's the one most reporting hides.

What you see	What's really happening	The control to apply
PMax ROAS far higher than Search or Shopping	PMax is absorbing high-intent brand searches	Add brand exclusions, run a dedicated brand Search campaign
Channel report shows heavy Search spend	Text and brand queries, not Shopping, are carrying the campaign	Test a feed-only build, tighten with negative keywords
Search terms you never targeted keep appearing	Final URL Expansion is matching loosely	Turn off Final URL Expansion on margin-sensitive catalogs
Blended ROAS strong, total revenue flat	Credit is shifting between campaigns, not growing the pie	Measure new-customer conversions, not blended return

The fix: exclude brand, then measure it honestly

Two moves solve most of this, and Google has quietly made both easier.

First, use brand exclusions. Unlike manual negative keywords, PMax brand exclusions now pull from Google's brand entity database, so they recognize your brand and its variants far more broadly than a keyword list ever could. You can hold up to 1,000 brands in an exclusion list, though most accounts only need 10 to 30 to cover their own name and a handful of direct competitors. If you need finer control, PMax now supports up to 10,000 campaign-level negative keywords, a big jump from the 100 cap that existed before March 2025. Use them to block the specific brand and competitor terms that slip through.

Second, and this is the step most brands skip, run a dedicated brand Search campaign. Excluding brand from PMax only helps if you then capture that traffic somewhere you can see and control. A tightly targeted brand Search campaign lets you value brand demand honestly, defend against competitors bidding on your name, and keep your cost per brand click low. Now your PMax campaign gets measured on what it should be measured on: incremental, non-brand conversions.

Set it up in this order

1. Build the brand Search campaign first, so brand traffic has a home before you close the PMax door.
2. Add brand exclusions to PMax, plus campaign-level negatives for competitor terms.
3. On margin-sensitive catalogs, switch off Final URL Expansion and automatically created assets.
4. Give it 40 to 60 days, then compare non-brand conversions and total revenue, not blended ROAS.

For catalog-heavy brands that want even tighter control, a feed-only Performance Max build is worth testing. It leans the campaign toward Shopping by using only your Merchant Center feed, with automatically created assets and Final URL Expansion switched off. Google doesn't recommend it, which is precisely why it's a useful lever for margin-sensitive accounts. Your product feed does most of the heavy lifting in any Shopping-driven campaign, so get that feed right before you blame the algorithm. That work sits where advertising meets site and feed optimization, and it pays off across every channel that reads your catalog.

What you give up, and when overlap isn't the problem

Being honest about this matters, because brand exclusions aren't free. When you carve brand out of PMax, the campaign loses some of its easiest conversions, and its measured ROAS will drop. That's the point. You're trading a flattering number for an accurate one. Expect a learning period, too. Any structural change resets part of PMax's optimization, and the system usually needs 40 to 60 days to settle. Don't judge the change in week one.

There's also a case for leaving things alone. If you're a newer brand with almost no brand search volume, cannibalization isn't really happening, and the extra structure just adds complexity you don't need yet. And PMax genuinely earns its keep on the discovery side: in one analysis of more than 180 ecommerce accounts, it drove around 23% higher conversion value on average than the campaigns it replaced. The goal isn't to distrust Performance Max. It's to stop letting it grade its own homework.

We've been running paid media for ecommerce brands since 2009, and the pattern holds across platforms. The brands that win with Google Ads in 2026 treat automation as an engine, not a strategy. They let PMax do what it's good at and wrap it in the controls and measurement that keep it honest, the same way they'd model channel mix or test incrementality anywhere else in the account. That's the difference between spending on Google and actually growing on it.

Where to start

Open your Search Terms report today and search for your own brand name. If it's driving a large share of PMax conversions, you've just found budget you can redeploy toward real growth this quarter. If you'd rather have a team run the audit, rebuild the campaign structure, and tie it to the rest of your Google and Meta spend, that's what our Google and Meta management is built to do. See what that's produced for other brands, or tell us what you're working with.