How brand exclusions work
Highly automated campaigns choose their own queries. Performance Max and the newer AI Max controls in Search decide what to match on, which means they will happily take searches for your own company name if nothing stops them. A brand exclusion is a list of brand names the campaign is then blocked from matching, whether that is your brand or somebody else’s.
The list is built once, in the shared part of the account, and applied to campaigns. Google matches variations of a listed name rather than only the exact spelling, so the list does not need every possible form, although it will not catch every misspelling a real person might type.
The scope matters more than the mechanism. Brand exclusions apply where there is a query to read, which means search and shopping inventory. Impressions on display and video surfaces have no search behind them, so a brand exclusion has nothing to act on there. There is also a mirror control, brand inclusions, which does the opposite and confines matching to the brands you name.
Why brand exclusions matter
Brand searches are cheap and they convert well, because the person already knows who you are. An automated campaign that is allowed to absorb them reports a strong return that it did not really earn, and that flattering figure then guides your budget decisions in the wrong direction.
Separating brand from everything else restores an honest comparison. Once Performance Max is kept off your own name, the results it reports are results from people who did not already know you, which is the only version worth judging it on.
Common mistakes with brand exclusions
The first is excluding brand from the automated campaign without making sure something else picks it up. If no search campaign is covering your brand terms properly, total brand volume falls and the loss is blamed on the exclusion rather than on the gap it exposed.
The second is expecting the split to be clean. Query matching is a judgement, not a filing system, and searches that mention your brand alongside a generic phrase sit awkwardly on the boundary. Some will still get through and some non-brand traffic may be caught.
The third is expecting exclusions to settle an argument about attribution. Removing brand from a campaign changes what that campaign reports; it does not tell you what would have happened to the sale if you had never advertised at all. Those are different questions and only a proper test answers the second.
How to act on it
Decide what you want to learn first. If the question is whether an automated campaign is finding new customers, exclude your own brand and watch its non-brand results. If the question is whether the campaign is worth running at all, the exclusion is only one part of a wider comparison.
Before switching it on, check your branded keyword coverage in Search so nothing is dropped, then measure brand volume across the whole account rather than campaign by campaign. Give it a full cycle before judging, and expect a period of instability while automated bidding adjusts to a narrower pool of queries.
