Performance Max Placement Exclusions

You open the Performance Max placement report after a flat week, and there it is: your ad ran on a toddler's colouring game, a celebrity-gossip aggregator, and something called dailytrendsnow. The obvious next question is how much each one cost you. You scroll for the spend column. There isn't one.

Short answer: Performance Max placement exclusions are set at the account level, or in a manager-account list that spans campaigns, and they cover URLs, apps, and YouTube channels. The report that feeds them shows impressions only, so treat exclusions as brand-safety hygiene rather than proof of where budget was wasted.

The takeaways

• PMax has no campaign-level placement exclusions by default. You exclude at the account level or through a manager-account (MCC) list that applies across campaigns.
• The placement report shows impressions, not clicks or conversions. Google's own help says to use it for brand safety, so you can see where an ad ran, never what it earned or cost.
• Made-for-Advertising sites are the target worth chasing. These run high-volume, low-intent inventory, and cutting them plus junk mobile-app categories is hygiene worth doing on a schedule.

What can you exclude in Performance Max?

More than the placement report implies, and most of it lives at the account level. You can add account-level placement exclusions for specific URLs, mobile apps, mobile app categories, and YouTube channels or videos. You can set a content-suitability tier (Limited, Standard, or Expanded inventory) that pulls your ads back from the edgier end of Google's inventory. You can exclude sensitive-content categories, apply campaign-level negative keywords, and use a brand list to keep branded search out of a given campaign. What you cannot do is exclude a single placement inside one campaign the way you would in an old Display campaign. The controls are blunter and sit a level up.

Why doesn't the placement report show conversions?

Because Google built it as a brand-safety view, not a performance view. When you try to add a conversion column, Google Ads flags the field as incompatible with Performance Max placements. The report lists placements with impressions and nothing downstream, and Google's help documentation states plainly that it should only be used to address brand-safety concerns (Google Ads Help). The reason is structural: PMax serves across Search, Shopping, Display, YouTube, Gmail, and Discover, and placements are only meaningful for some of that inventory. So the report is partial by design. It tells you where an ad appeared, and it stays silent on whether that appearance did anything for you.

Where do you set the exclusions?

At the account level, and once for the whole account. In Google Ads you build a placement exclusion list under account-level settings, then it applies to eligible campaigns underneath it. If you manage several accounts, a manager-account exclusion list lets you push one blocklist down to all of them, which is how most agencies run it. The placement report itself lives under Campaigns, in the Insights and reports menu, through the report editor. Worth knowing before you start: because the exclusion is account-wide, you are making a decision for every PMax campaign at once, not tuning one. That is fine for genuine junk. It is a reason to be conservative with anything borderline.

How do you spot junk before it drains budget?

Read the placement report for concentration rather than precision. Sort by impressions and look for the pattern that gives Made-for-Advertising sites away: a long tail of odd domains and low-quality apps soaking up a surprising share of your impressions. Kids' games, wallpaper apps, and auto-refreshing news clones are the usual suspects. Maintained MFA blocklists circulate publicly and are a reasonable starting seed, so you are not building the list from scratch. Then set content suitability to Limited inventory if brand risk matters more than raw reach. The honest caveat sits underneath all of this: you are acting on an impressions signal, so you are cleaning up where your ads show, and inferring the budget effect rather than measuring it.

So are placement exclusions worth the effort?

Yes, as hygiene, with the right expectation. Excluding MFA sites and toddler apps keeps your brand off inventory you would never buy on purpose, and it nudges the algorithm toward channels that tend to convert. What it will not give you is a clean before-and-after on spend, because the data to prove that was never in the report. This is the same discipline that runs through every good testing setup: judge a decision on a number you can trust, and be honest about the ones you are only inferring. That is the whole idea behind ad intelligence as we build it, and it is why the composite score we use weights measured ROAS and revenue per install over a proxy the platform hands you for free. If you run Meta as well, the Audience Network version of this exact problem is worth a look in the Meta placement audit.