Microsoft Ads Excluded Content Terms: 1,000 Page-Title Terms

Microsoft Advertising now lists Excluded Content Terms as a standard brand-safety control. It lets an advertiser block ads from appearing next to web pages whose titles contain terms on a list, up to 1,000 of them. When we read Microsoft’s content suitability help page on September 28, 2026, the page no longer carried the notice that not every account had the feature yet. An earlier copy of the same page on Microsoft Learn, last updated on July 27, still shows it.

We found no Microsoft announcement of the change, so the help page is the record. It answers the practical questions: which campaigns the control covers, how it fits with the two older controls, and one behaviour that surprises people who manage several clients or brands in one account.

Microsoft’s content suitability help page gives a simple example. An advertiser selling tech devices adds the term “faulty tech devices,” and Microsoft stops showing its ads next to page titles that mention them. The page does not say whether matching is exact, by phrase or by close variant, so test a few terms before relying on a long list.

Trade press reported the control for Audience Ads in August as a limited release. The difference now is the missing availability notice, which suggests every account should see the setting. The rollout may not be complete, though: when we re-checked on September 29, Microsoft’s Performance Max setup page still said not everyone has excluded content terms yet. If yours does not, the setting sits under Tools, then Content suitability.

Underneath all three sits a default. The help page says IAS checks sites that carry audience placements for eight kinds of content that are never suitable for advertising, including adult content, hate speech and violence, and Microsoft blocks ads on sites IAS rates as medium or high risk. Microsoft also runs its own checks on the sites it owns.

The new control fills a gap the other two leave. A category is too broad for a risk that only one brand cares about, and a site list cannot catch a single bad article on an otherwise good site. A page-title term can, for example the name of a recalled product or a competitor’s scandal.

For a single-brand account that is harmless, even helpful. For an agency or a company that runs several brands or product lines in one account, it is a trap. A term that protects one brand can cut reach for another that sells next to exactly that content. A retailer that excludes “recall” for its baby products, for example, would also block its car-parts campaigns from pages about car recalls. The simplest defence is to keep brands with different risk profiles in separate accounts, and to review the account list after anyone adds terms during campaign setup.

A thousand terms is a lot, and more is not safer. Every term removes placements, and the help page does not mention any report of how many placements a term has blocked. Three habits keep a list useful:

• Start from real incidents. Use placement reports and past complaints to find the page titles you actually objected to, and write terms for those.
• Prefer phrases to single words. “Faulty tech devices” blocks a narrow set of pages; “faulty” alone could block reviews and repair guides your buyers read.
• Review it on a schedule. Terms tied to a news event stop being useful once the story fades, but they keep removing placements until someone deletes them.

Microsoft’s other recent changes give its automated bidding more room, including the removal of the max CPC setting on October 1. A term list keeps one placement limit in the advertiser’s hands as bidding becomes more automated. Google’s equivalent debate is covered in our post on AI Max text guidelines and brand safety.