Methodology

Data source

All data comes from the Meta Ad Library under the EU Digital Services Act transparency provisions. For every ad delivered to the EU, Meta publishes its creative, delivery dates, targeting summary, and — crucially — eu_total_reach: the cumulative number of EU accounts-in-reach over the ad's whole run, broken down by country, age, and gender.

We snapshot tracked brands daily. Meta publishes reach as a running total with no time axis; the daily snapshots are what create one. The difference between yesterday's and today's total is the ad's reach velocity — that's how we see momentum and fatigue, not just size.

Why reach is “verified” — and spend never is

Reach is verified: it is a field Meta publishes, and we report it as published. Spend is not published for commercial EU ads — Meta discloses spend only for political and issue ads. Any tool showing you competitor spend for commercial ads is estimating, whether it says so or not. We say so, and we never attach the word “verified” to a spend figure.

When we combine reach across the ads that share one creative, the number is always qualified “across N placements”: the same person can be reached by two placements, so a combined figure is an upper bound.

Spend estimates

Estimated spend = verified EU reach × a CPM benchmark range for the relevant country and month. Benchmarks are calibrated from real EU campaign data and public sources, and every estimate is shown as a range (low–high), never a point figure. Where no benchmark exists for a market, we show no estimate rather than a bad one.

Winner and killed-test classification

Deterministic rules over delivery data, applied per brand:

• Winner — active for 21+ days and cumulative reach at or above the brand's own 75th percentile. Brands scale what works; longevity plus relative reach is the signal.
• Killed test — stopped in under 7 days and below the brand's 25th percentile. Fast kills are decisions, and they're often more informative than the winners.
• Everything else is simply running.

What we can't know (and won't pretend to)

• US-only campaigns are invisible. The DSA data covers EU-delivered ads only. If a brand spends exclusively outside the EU, it produces zero rows. Every report states “EU-delivered ads.”
• Payer/beneficiary is self-declared. We use it to map agency relationships, and we label it as a declared signal, not ground truth.
• Catalog ads are excluded. Dynamic product-feed ads (hundreds of near-identical variants from one template) are detected and excluded from creative analysis — they'd otherwise drown the actual creative work.
• Reach is accounts-in-reach, not people. Meta's metric, reported as Meta defines it.

Corrections

If a published number doesn't hold up to this page, we correct the post and say we did. Email hello@creativescout.app.