How to Stop Performance Max From Cannibalizing Your Brand Traffic

PMax eats brand searches because they are cheap and convert. Here is how to fence it off without killing performance.

· QueryCut

The short version

  • PMax takes brand traffic because it is the cheapest, highest-converting inventory in your account, not because of a setting you forgot.
  • Apply a brand negative list plus request rep-level brand exclusions, since self-serve PMax brand controls miss misspellings and sub-brands.
  • Stand up a high-impression-share brand Search campaign before you exclude brand from PMax, never after.
  • Find leaked variants by diffing your brand Search query report against PMax's brand-themed search category conversions.
  • Verify with impression share up, PMax brand conversions down, and total conversions flat or better over 7 to 10 days.

To stop Performance Max from cannibalizing your brand traffic, apply an account-level brand negative keyword list to PMax (or ask your Google rep to add brand exclusions, since the self-serve list has limits), and run a separate exact and phrase match brand Search campaign to absorb those queries at a controlled CPC. PMax gravitates to brand searches because they are the cheapest, highest-converting inventory in your account, so it will claim them unless you explicitly wall it off. The rest of this post explains why the leak happens, the exact steps to close it, and how to confirm it actually closed.

Why PMax steals brand traffic in the first place

Performance Max optimizes toward a goal (conversions or conversion value) across every channel Google owns, and it has no concept of "this query belongs to my Search campaign." When someone searches your brand name, that auction is the single most attractive one in your account: high intent, high conversion rate, low CPC. If PMax can serve on it, the algorithm will, because doing so inflates the campaign's reported ROAS at almost no cost.

This creates a reporting illusion. PMax looks like a star performer partly because it is harvesting demand you already created through your brand, your other campaigns, your email, and word of mouth. That conversion was going to happen anyway. PMax just stepped in front of it and took credit, and in the process it drains budget and attribution away from campaigns you can actually control.

The mechanism matters because it predicts behavior you have not seen yet: any time PMax has a path to a cheap, high-intent query, it will take it. Brand is the obvious case, but the same logic applies to competitor-adjacent or navigational queries where your site is the clear answer. Fence the brand case first, then watch for the pattern elsewhere.

The fix, in order of priority

Do these three things today. The first two stop the leak; the third makes sure the traffic lands somewhere good.

Step one: add a brand negative keyword list to PMax. In the account, go to Tools > Shared library > Negative keyword lists, build a list with every brand variant and common misspelling, then apply it. For PMax specifically, account-level negatives can be applied in Admin > Account settings > Negative keywords, or you can have your Google representative add brand exclusions directly to the PMax campaign. Self-serve brand exclusions in PMax rely on Google's brand lists, which do not catch every misspelling or sub-brand, so the shared negative list plus a rep request together give you better coverage than either alone.

Step two: build or strengthen a dedicated brand Search campaign. Use exact and phrase match on your brand terms, a tight budget, and manual or target impression share bidding aimed high (90 percent or above) so you reliably win your own name at a low CPC. This campaign becomes the designated home for brand queries. Without it, excluding PMax from brand just hands those searches to competitors or to organic, which may or may not be what you want.

Step three: exclude brand from your generic Search campaigns too, so you can read each campaign cleanly. The goal is one campaign that owns brand and everyone else blocked from it. Once traffic is sorted by intent, your ROAS numbers finally mean something.

Finding the brand variants PMax is actually leaking

The hard part is not blocking brand, it is finding every variant PMax is capturing, because PMax does not give you a full search terms report. You get a limited search categories view under the campaign's Insights, which groups queries into themes and hides the raw terms. That means you cannot just read off the leaked brand searches the way you can in a Search campaign.

Use a diff method. Pull the query-level data from your brand Search campaign (Campaigns > Insights and reports > Search terms) and list every brand query that drives conversions there. Then open the PMax campaign's Insights > Search categories and find the brand-themed category and its conversion count. If PMax is reporting brand-themed conversions that your Search campaign is not capturing, the gap is your leak: those are brand searches PMax is intercepting before your Search campaign ever sees them. The specific misspellings often show up in the search categories detail or in your Search campaign's lost-to-negative data once you start blocking.

Also mine Google Search Console and your site's internal search for how people actually type your brand. Hyphenated forms, spaced-out forms, your brand plus "login" or "reviews," and phonetic misspellings are the variants that slip past a tidy negative list. Add each one to the shared list as you find it.

The failure mode: do not block brand blind

The trap practitioners fall into: they see PMax reporting great numbers, assume it is all incremental, rip brand out of PMax, and then watch total account conversions dip. Here is why. Some of those brand searches were genuinely new users who found you through a PMax display or YouTube touch earlier in their journey. Blocking brand in PMax does not always redirect that conversion cleanly to your Search campaign, especially if your brand Search bids or budget are too low to win the auction.

So sequence it. Stand up the brand Search campaign with high impression share bidding first, confirm it is winning your brand terms, and only then exclude brand from PMax. If you flip the order, there is a window where nobody is defending your brand name and a competitor bidding on it can poach clicks at your expense.

Watch out for a second trap: a shared negative list that is too aggressive. If you add a broad token like your brand's common word (say your brand is "Summit" and you block "summit"), you will also block legitimate generic searches. Keep brand negatives as phrase or exact match on the full brand string, not single common words in broad match.

How to verify the leak actually closed

After you apply the negatives, do not assume it worked. Changes to PMax exclusions can take 24 to 48 hours to fully propagate, and the search categories report lags further. Give it a few days of data before judging.

Check three things. First, your brand Search campaign's impression share (add the Search impression share column) should rise after you grab the traffic PMax was taking. Second, the PMax brand-themed search category's conversion count should fall over the following week. Third, total account conversions and conversion value should hold steady or improve. If total conversions drop meaningfully while PMax brand conversions fall, that is your signal that some of the brand traffic was more incremental than you thought, and you may want to raise brand Search bids rather than leave those auctions uncontested.

Set a simple decision rule: if after 7 to 10 days your brand Search campaign is winning 90 percent-plus impression share and total conversions are flat or up, the fix worked and you are now reading clean numbers. If total conversions fell, revisit your brand Search bids and budget before blaming the exclusion.

Questions people also ask

Can I see the exact search terms PMax matched my brand on?

Not fully. PMax only shows grouped search categories under the campaign's Insights, not a raw search terms report. You infer leaked brand variants by comparing those brand-themed categories against the query-level data in your brand Search campaign and by mining Search Console for how people actually type your brand.

Do PMax brand exclusions work instantly?

No. Applied negatives and brand exclusions can take 24 to 48 hours to propagate, and the search categories reporting lags further. Wait several days of data before deciding whether the leak closed, and judge it over a 7 to 10 day window.

Should I ever let PMax keep brand traffic?

Sometimes. If some brand searches are genuinely new users PMax introduced upstream through display or YouTube, blocking brand can reduce incremental conversions. The safe approach is to control brand in a dedicated Search campaign first, then exclude it from PMax and watch whether total account conversions hold.

Why not just add brand as a negative in every campaign and be done?

Because excluding brand from PMax without a brand Search campaign winning those auctions leaves your brand name open for competitors to bid on. Block brand everywhere except one dedicated brand Search campaign set to high impression share, so the traffic has a controlled, cheap place to land.

What is the single fastest safe action I can take right now?

Confirm your brand Search campaign is live with exact and phrase match brand terms and target impression share of 90 percent-plus. Once it is winning your brand name, apply an account-level brand negative list and request brand exclusions from your Google rep for the PMax campaign.

More from the blog

Landing Page Experience: The Changes That Actually Move ItImpression Share Lost to Rank vs Budget: What Each One MeansHow Many Campaigns Is Too Many for the Budget You Have?Primary vs Secondary Conversion Actions: What Each One Actually Changes