What Advertisng’s public-ad corpus can—and cannot—tell an agency
As of August 9, 2026, the Advertisng research corpus contained 6,957 public ad records across 541 advertisers: 6,082 Meta records and 875 Google records. This is a product-corpus snapshot, not a market-share estimate.
Reviewed August 9, 2026 by Advertisng Research Team
The corpus at a glance
- • 6,957 total public ad records in the database snapshot
- • 6,082 Meta records (87.4% of the snapshot)
- • 875 Google records (12.6% of the snapshot)
- • 541 advertiser records
- • 6,356 records marked active by the current ingestion state
How to interpret the numbers
The platform split reflects Advertisng’s present source coverage and ingestion priorities. It does not measure total advertising activity on Meta or Google. “Active” is a source and ingestion-state signal; it does not establish current delivery, spend, conversions, or profitability.
Three useful agency applications
- • Build a category map of recurring offers, hooks, formats, and proof devices.
- • Compare how the same advertiser expresses a proposition across Meta and Google public surfaces.
- • Create research-backed hypotheses for briefs, then validate those hypotheses with the client’s first-party campaign results.
What this snapshot cannot prove
- • Which ad generated the most revenue or profit
- • Audience targeting, bids, budgets, attribution settings, or conversion rate
- • Whether repeated creative is a controlled test, a localization, or an operational duplicate
- • Market share, platform-wide prevalence, or causal performance relationships
Reproducibility note
Counts were produced from the live Advertisng production database on August 9, 2026, grouped by stored platform. The next quarterly edition should preserve this baseline, state any ingestion or classification changes, and compare only like-for-like fields.