Is Performance Max Cannibalizing My Search Campaigns?

Performance Max is cannibalizing your Search campaigns if branded Search impression share has fallen without a budget change, PMax conversions are rising while total account conversions stay flat, and Search cost per acquisition is climbing. Those three signals together mean PMax is winning auctions your Search campaigns were already winning more cheaply, and booking the conversion as its own. Brand exclusions fix it in about ten minutes, and branded impression share usually recovers inside two weeks.

What does Performance Max cannibalization actually mean?

Cannibalization is not PMax performing badly. It is PMax performing well on demand you already had. Someone who already knows your company types your brand name into Google. That person was going to convert. Before PMax existed they clicked your brand Search ad at maybe forty cents, or your organic listing at nothing. Now Performance Max is eligible for that same auction, wins it, and records the sale in the PMax column.

Nothing about the customer changed. What changed is which line item takes credit, and what you paid. The PMax dashboard shows a cheap conversion and a healthy return on ad spend, so the campaign looks like your best performer, and the natural response is to give it more budget. That is the trap: you fund expansion out of a number that was never incremental in the first place.

Published analyses put the size of it at 15% to 30% of inflated apparent ROAS and 8% to 15% of total PMax budget spent on brand demand. On a $20,000 monthly PMax budget that is $1,600 to $3,000 a month buying customers who already had your name in their head.

How do I know if Performance Max is stealing my branded search?

Four checks, each about fifteen minutes. Run them in this order, because the first one is the cleanest evidence and the last one is the hardest to argue with.

Check 1: branded Search impression share over 90 days

Open your brand Search campaign and add the Search impression share column, then chart it daily for the last 90 days. If that line steps down and the drop lines up with the date a Performance Max campaign launched or had its budget raised, you have your answer. Nothing else moves branded impression share that abruptly. Competitors bidding on your name produce a slow slide, not a step.

Look at absolute impressions on the brand campaign too. Cannibalization often shows there before impression share, because share is a ratio and both halves are moving.

Check 2: total account conversions, not campaign conversions

This is the check that settles arguments. Pull total conversions for the whole account, month over month, alongside PMax conversions for the same period.

Pattern	PMax conversions	Account total conversions	What it means
Genuine incrementality	Up 120	Up 110 or so	PMax found new demand. Fund it.
Partial cannibalization	Up 120	Up 40	Roughly two thirds moved from other campaigns.
Full cannibalization	Up 120	Flat	Nothing was gained. Money changed pockets.
Negative	Up 120	Down	PMax is displacing cheaper conversions and costing more.

Most accounts land in the second row. The uncomfortable part is that partial cannibalization still looks like success in every per-campaign report you will ever be shown.

Check 3: Search campaign CPA drift

When PMax takes the easy conversions, your Search campaigns are left with the hard ones, and their cost per acquisition rises even though nothing about them changed. Compare Search CPA in the 60 days before PMax launched with the 60 days after. A rise of 20% or more in a Search campaign whose keywords, budget, and landing pages are untouched is a strong secondary signal.

This one is easy to misread, so check the obvious alternatives first: seasonality, a competitor entering the auction, a landing page change, a tracking change. If none of those apply, PMax is the remaining explanation.

Check 4: how much of PMax is brand

Google gives you less here than it should, but two routes work. In the PMax campaign, open Insights and look at the search categories report, which groups the queries PMax served against. If your company name or close variants sit near the top by conversions, that is the cannibalization, visible directly.

The second route is a script or the Google Ads API pulling search term insights per campaign, which is more complete but needs setup. If neither is available, the indirect version works well enough: apply brand exclusions for four weeks and measure what happens to total account conversions. If they hold steady while PMax conversions drop, everything that disappeared was brand.

How do I stop Performance Max from bidding on my brand?

Set brand exclusions account-wide, then make sure a brand Search campaign is live to catch the queries PMax is now blocked from. In the Google Ads interface go to Tools, then Shared library, then Brand lists, and create a list containing your company name and its common misspellings. Applying it at the shared-library level is the important part, because per-campaign exclusions get forgotten the next time someone spins up a new PMax campaign.

Once the list exists, PMax is blocked from auctions where Google detects your brand. Brand exclusions have been available since 2023 and are now the default recommendation for any advertiser with existing brand demand, which is nearly everyone who has been trading for more than a year.

Then check the brand Search campaign is actually able to catch what comes back. It needs enough budget headroom to absorb the returning volume, exact and phrase match coverage of your name and its variants, and a bid strategy that will not throttle at the low cost per click brand terms deserve. Applying exclusions without a brand campaign ready just hands the traffic to whoever else is bidding on your name.

The variant worth considering: split brand and non-brand PMax

Excluding brand entirely is the right call for most accounts. There is a second approach for large ecommerce advertisers: run one PMax campaign with brand excluded and a separate brand-only PMax with its own budget and its own target. It gives you a clean read on each, at the cost of more campaigns to maintain. If you are unsure, exclude first. You can always split later, and the split is only worth the complexity when brand volume is large enough to justify its own budget line.

How long does it take for Search to recover after brand exclusions?

Branded Search impression share typically recovers within one to two weeks. Blended cost per lead takes longer, usually three to four weeks, because Performance Max has to re-learn where its conversions come from once the cheapest ones are off the table. Use a four-week minimum test window and do not judge anything in week one.

Expect the PMax dashboard to look worse during that window, and expect that to feel wrong. It is not. PMax conversions fall because the brand conversions moved back to the brand campaign where they cost less, and PMax ROAS falls because the number was partly borrowed. What should hold is total account conversions. Here is what a normal recovery looks like on a mid-sized account.

Metric	Week 0 (before)	Week 2	Week 4	Reading
Branded Search impression share	52%	86%	91%	Recovers fast, as expected
PMax conversions	310	225	240	Drops, then partly rebuilds on new demand
PMax reported ROAS	4.1x	2.9x	3.1x	Falls to something honest
Total account conversions	640	631	668	Holds, then improves. This is the number that matters
Blended cost per acquisition	$61	$58	$53	Falls, because you stopped paying twice

If total account conversions fall and stay down after four weeks, brand was not the problem and something else changed at the same time. Go back and check tracking before reversing the exclusions.

Does Performance Max cannibalize organic traffic too?

Yes, and it is harder to see because the loss shows up in a different system. When PMax serves an ad on your brand query, it sits above your own organic listing, and a share of the people who would have clicked the free result click the paid one instead. Your Search Console clicks drop, your Google Ads clicks rise, and the total is roughly unchanged except you are now paying for part of it.

The way to catch this is to look at brand query clicks in Search Console over the same window you are testing exclusions in. If organic brand clicks rise when PMax brand spend stops, that is the overlap, and it was costing you real money for traffic that was already free. Keeping a live view of paid and organic performance for the same queries next to actual revenue is the only way to see it clearly, and it is one of the arguments for having every channel and your store reporting into one dashboard rather than judging each platform on its own numbers.

What if I still cannot tell?

Run the exclusion as an experiment rather than a decision. Apply brand exclusions, hold every other variable still for four weeks, and write down total account conversions and blended cost per acquisition before you start. If those two numbers hold or improve, the brand spend was not incremental and you have just saved 8% to 15% of your PMax budget. If they genuinely fall, reverse it and you have lost one month of a slightly cheaper conversion mix.

That asymmetry is why this is worth doing even when you are unsure. The downside is one month of a marginally worse blend; the upside is finding out that a meaningful slice of your best-looking campaign was buying customers you already owned.

Cannibalization is also rarely the only thing wrong in an account where it is happening, because both problems come from the same cause: nobody has looked closely in a while. The step-by-step Google Ads audit covers the full check in dependency order, and the guide to finding wasted spend in Google Ads covers the rest of where budget leaks. If the account spans more than one platform, a multi-channel PPC audit adds the cross-platform double counting that a Google-only review structurally cannot see.

Frequently asked questions

Should I exclude my brand from Performance Max?

Yes for almost any advertiser with existing brand demand. Set the exclusion account-wide under Tools, Shared library, Brand lists, and keep a brand Search campaign live to catch the returning queries. The exception is a brand-new company with no brand searches at all, where the exclusion has nothing to block and costs you nothing either way.

Why did my Search impression share drop after launching Performance Max?

Because PMax became eligible for the same auctions and won some of them. Google does not run both your campaigns against each other in the way advertisers assume; the campaign with the higher Ad Rank serves, and on brand queries PMax often has it. The impression share your Search campaign lost went to your own PMax campaign, not a competitor.

Does Performance Max convert better than Search?

On paper often yes, in reality frequently no. An Adalysis study across roughly 3,300 campaigns found Search converted at a higher rate than Performance Max when both were eligible for the same query. Reported PMax numbers look stronger largely because the campaign is absorbing high-intent brand traffic that would have converted anywhere.

How much of my budget should Performance Max take?

There is no single right split, but PMax should be funded on top of a healthy Search account rather than out of it. Ecommerce accounts with a large, well-structured product feed justify a higher share because the feed does the targeting. Lead generation accounts usually do better keeping the majority in Search, where a bad query can be excluded by name. Our Performance Max optimization page covers how to decide, and automated Google Ads management covers having the split watched daily rather than at a monthly call.

Will brand exclusions hurt my Performance Max performance?

The reported numbers will drop and the real ones should not. PMax conversions and PMax ROAS both fall because the cheapest conversions were brand. Total account conversions and blended cost per acquisition are the pair to watch, and on most accounts blended CPA improves within four weeks because you stop buying the same customer twice.