Skip to content

Problem

Performance Max spending on brand searches: how to fix it

Performance Max will serve on searches for your own name unless you tell it not to, and those searches convert at a rate no prospecting can match. The result is a campaign that reports a strong return while buying orders you were already going to receive. The repair is to measure how much of the spend is brand, decide deliberately whether you want to pay for it, and then use brand exclusions rather than hoping the algorithm redirects itself.

By CartKernel · Last reviewed 2026-09-07

Does this look familiar?

  • Reported return on ad spend is far higher than the account has ever produced elsewhere
  • Organic and direct orders fell while Performance Max orders rose by a similar amount
  • Total revenue is flat while paid revenue looks like it grew
  • The search terms insights view is dominated by your own name and product names
  • A brand search campaign lost impression share the week Performance Max launched
  • Conversions arrive with very short click to purchase times and almost no assisted path

Causes, ranked

Why it happens, most common first

Check them in this order. The first two account for most cases we open.

  • most common

    No brand exclusion has been applied to the campaign

    By default the campaign can match on any query the system believes will convert, and your own name is the easiest of those. Brand exclusions are a specific setting that has to be created as a brand list and applied to the campaign. Without it, brand traffic is simply the cheapest conversion available.

  • most common

    Audience signals and search themes point back at existing customers

    Uploading a customer list, using site visitor segments or writing search themes that include your brand name tells the system where the good conversions are. It obliges. The signals are meant to accelerate learning, but they read as a request to buy people who already know you.

  • common

    The bid target rewards easy conversions

    A high target return can only be reached on high intent traffic. Setting an ambitious target pushes the campaign toward the searches most likely to convert, which are almost always brand, returning visitor and remarketing placements rather than new demand.

  • common

    The feed carries brand terms in every title

    When each product title starts with the store name, the campaign has an obvious reason to match brand queries. This is common for private label catalogues where the brand and the store are the same word, and it makes brand and generic traffic hard to tell apart.

  • occasional

    A separate brand search campaign was paused or unfunded

    If the dedicated brand campaign stops covering those queries, Performance Max steps into the gap. The spend does not disappear, it moves, and it becomes harder to see because Performance Max reporting groups it with everything else.

  • occasional

    New customer acquisition settings are switched off

    The campaign can be told to value new customers more highly, or to bid only for them. When neither is configured, a returning customer purchase counts exactly the same as a first order, so nothing steers the campaign toward new demand.

The fix

In this order

Each step is something you can do today. Do them in sequence; skipping ahead is how a review fails twice.

Prevent it next time

  • Apply a brand exclusion at launch rather than after the first strong report
  • Report new customer orders and blended acquisition cost next to campaign return
  • Keep brand defence in its own campaign so its cost is always visible
  • Review search terms insights monthly and add the queries that should never cost money
  1. Measure the brand share before changing anything

    Open the search terms insights for the campaign and categorise the terms into brand and non brand. Note the share of conversions and, where the view allows, the share of clicks. Write the figures down so the effect of the change is measurable later.

  2. Compare paid growth against total revenue

    Line up Performance Max revenue against total store revenue for the same weeks. If total revenue is flat while paid revenue rose, the campaign is claiming orders rather than creating them. This is the number that decides whether the change is worth making.

  3. Create a brand list and apply it as an exclusion

    Build a brand list containing your store name, its common misspellings and your own product line names, then apply it to the campaign as a brand exclusion. Give it a day or two to take effect before reading results.

  4. Add campaign level negative keywords for the remainder

    Brand exclusions cover the brand itself. Add negative keywords for support queries, tracking queries, login and returns phrases that convert badly and would otherwise absorb budget once brand is removed.

  5. Rewrite audience signals toward new demand

    Remove customer lists and site visitor segments from the audience signal, and replace search themes that name your brand with the category and problem language a new customer would use. Signals guide rather than restrict, so make them describe the person you do not have yet.

  6. Run a dedicated brand search campaign alongside it

    Keep brand coverage in a separate search campaign where you can see its cost and its conversion volume on their own. That way the decision to defend the brand is explicit and priced, rather than hidden inside a blended report.

  7. Reset the target for the traffic that remains

    Once brand is out, the campaign faces harder auctions and the achievable return falls. Lower the target to something the non brand traffic can reach, in one step, and allow the learning period to complete before judging the result.

  8. Re-measure against the baseline you recorded

    After a few weeks, compare total store revenue, new customer orders and blended acquisition cost with the figures from step one. Judge the change on those, not on the campaign level return, which will look worse by design.

When to get help

Bring in help when the honest answer to whether the campaign is incremental is not obvious from the reports, which is usually the case for stores with strong organic and direct demand. Separating brand from generic performance properly means holdout tests, blended reporting across channels, and a view of contribution rather than reported return. It is also worth help when brand exclusions have been applied and the campaign has not stabilised, because the rebuild involves feed structure and campaign splits rather than a single setting.

Get a Growth Analysis

Free. We reply within one business day.

Questions

Asked alongside this problem

By CartKernel · Last reviewed

Do brand exclusions remove all brand traffic from Performance Max?

They remove serving on search and shopping inventory for the brand terms you list, which is where most of the brand spend sits. Some remarketing and display placements can still reach people familiar with you. Expect a large reduction rather than a clean zero, and check the search terms view afterwards.

Should I stop bidding on my brand entirely?

That depends on whether others advertise on your name and how much of your traffic is already loyal. Turning brand off for a defined period and watching total revenue rather than paid revenue is the only reliable way to find out for your store. Many stores find some defence is worth it and full coverage is not.

Why did my return on ad spend fall after excluding brand?

Because the easiest conversions were removed from the numerator while the harder ones remained. That is the intended outcome. The figure to watch is whether total store revenue held while the campaign now buys traffic that was not already coming, which is what the campaign was funded to do.

Can I see the exact search terms Performance Max is matching?

The reporting gives search term insights grouped into categories rather than the full term by term list available for search campaigns. It is enough to judge the brand share and spot obvious waste. For finer control, pair it with a search campaign where the term report is complete.

Related problems, answers and terms

All problems
ProblemPerformance Max not converting: how to fix itPerformance Max not converting usually means the campaign is starved of signal, assets or eligible products. Diagnose in order before restructuring.OpenProblemCompetitors bidding on your brand name: how to fix itOther advertisers appearing on your brand name is allowed in most cases. What you can control, what you can report, and how to hold the position.OpenProblemGoogle Ads ROAS dropped: how to fix itWhen Google Ads ROAS dropped, decide first whether revenue fell or measurement changed. The order of these checks saves weeks of wrong fixes.OpenProblemShopping CPCs rising: how to fix itShopping CPCs rising is often a mix shift rather than a price rise. Separate the query, product and auction causes before changing any bid.OpenAnswerShould an ecommerce brand bid on its own brand name?Most stores should bid on their own brand name, for defence and control rather than for the reported ROAS. How to decide and how to test it honestly.OpenAnswerHow do you set new customer acquisition goals in Google Ads?Setting a new customer acquisition goal in Google Ads means telling bidding what a first order is worth. How to derive that number and configure the goal.OpenGlossaryIncrementalityIncrementality is the share of conversions that only happened because of the ad. Formula, a worked geo holdout, and how to run the test properly.OpenGlossaryReturn on ad spend (ROAS)Return on ad spend (ROAS) is attributed revenue divided by ad spend on one platform. Formula, worked example, why it matters and where it misleads.Open

Stop the leak.

A free Growth Analysis finds what is broken and ranks it by the revenue it costs you.