Google removed 3.4B ads and 5.6M+ advertiser accounts in 2021, after implementing a three-strikes rule for deceitful practices, dangerous products, and more
Context & Ripple Effects
Google's annual ads-safety reports have become the industry's yardstick for platform integrity since it first disclosed removing 3.2B bad ads in 2017. The 2018 report showed removals dipping to 2.3B alongside roughly 1M terminated advertiser accounts [[a:939550]], so the 2021 figures mark a return to higher ad volumes with a structural change underneath.
What makes the 2021 report matter is the three-strikes rule: enforcement shifts from pulling individual ads to penalizing repeat-violator accounts, which is why account terminations jumped from about 1M in 2018 to over 5.6M. That account-level framing is what the later reports build on.
First-order effects
- Advertisers running deceitful-practice or dangerous-product campaigns now face graduated account-level penalties under the three-strikes rule instead of one-off ad removals.
- The 5.6M+ account terminations are a step change from the ~1M bad advertiser accounts Google reported terminating for 2018, meaning far more businesses lose ad access entirely rather than just losing individual creatives.
Second-order effects
- Enforcement volume keeps compounding after this report — by 2023 Google was blocking or removing 5.5B ads and suspending 12.7M advertiser accounts [[a:850876]] — making policy compliance a recurring operating cost for anyone buying Google ads at scale.
- Publishers and app developers, already subject to site- and app-level bans in earlier reports, face the same ratchet as account-level strikes extend beyond advertisers to the supply side.
Third-order effects
- The account-strike model Google adopted here is the precursor to its later AI-driven enforcement: by 2025 it reported a record 8.3B ads blocked while suspending 36% fewer advertisers, crediting AI tools [[a:1167148]] — suggesting automated systems, not human reviewers, increasingly decide who may advertise at all.
The trend: Ad-platform enforcement is shifting from removing individual ads to mass account suspension governed by strike systems, with AI progressively taking over both detection and adjudication.