What every platform ad library actually covers

Competitor ad monitoring fails in a predictable way. Someone checks four ad libraries, finds ads in two of them, and concludes the competitor advertises on two platforms. The real conclusion available from that work is that two libraries had coverage for the countries checked on the day checked, which is a much smaller claim.

This page is the coverage map we work from. It is worth having in front of you whether you monitor by hand, build a collector, or buy a tool, because every vendor in this category is constrained by the same boundaries and most describe them in aggregate.

Sources with a usable public library

• Meta (Facebook, Instagram). Global commercial coverage, and the only large library that is genuinely worldwide. The catch is retention: commercial ads are present while active and gone once stopped. Ads delivered in the EU, UK and Brazil carry a published reach figure in the per-ad details panel. Detail in the Facebook Ad Library.
• Google and YouTube. Global, verified advertiser identities, Search, Display, Shopping and video in one view, with an explicit anywhere-region option. Terms restrict use to internal information and research. Detail in the Ads Transparency Center.
• TikTok. Commercial Content Library, scoped to ads available in the EEA, Switzerland and the UK. Not worldwide, and not the Creative Center. Detail in TikTok's two libraries.
• Microsoft. Anonymous Ad Library covering EEA delivery. In our own runs it caps out around a thousand total results, which is a source cap rather than exhaustion, and the difference matters when you are deciding whether a competitor has more ads than you can see.
• Snapchat. Documented Ads Gallery covering EU delivery over the previous twelve months.
• Pinterest. Currently 30 countries in its own selector: EU27 plus Brazil, Norway and Türkiye.
• X. A portal search covering its EU27 selector. We treat it as unvalidated until a live ad row confirms the field contract, so it stays switched off rather than shipping thin.

Sources without one

• LinkedIn. A good public library with about a year of retention, and no programmatic access. Monitoring is manual unless a vendor holds an approved product grant. Detail in the LinkedIn Ad Library.
• Reddit and Amazon. As of September 2026 we found no exhaustive first-party competitor transparency library for either. A tool claiming coverage here is claiming something we could not verify, and it is a fair question to put to them.

Read that list again with geography in mind. Meta and Google are global; TikTok, Microsoft, Snapchat, Pinterest and X are European-scoped, because the Digital Services Act is what made those repositories exist. So a cross-platform ad count is a comparison between a worldwide number and a regional one unless someone has normalised the geography, and almost nobody says whether they have.

What no commercial ad library publishes

• Spend. Political and election ads carry ranges where regulation requires it. Commercial ads carry nothing. Every commercial spend figure in every tool is modelled from panel data or inference.
• Impressions or conversions. EU reach disclosures are a real observation and are neither of these, and EU, UK and Brazil figures measure different regions, so summing or trending them together produces a number with no meaning.
• Whether the ad worked. Duration is the closest honest proxy. An ad still running after four months is an ad someone kept funding, and that is a stronger signal than most modelled metrics.

The two questions that decide whether monitoring works

Did you resolve the right advertiser? Brands run multiple pages, verified entities and acquired-brand accounts. Getting this wrong produces a confident, wrong answer that nothing downstream can detect, which makes it the worst failure mode in the category. Confirm the identity per platform before collecting anything.

Was the collection complete? An incomplete run and a stopped ad look identical. If a process cannot distinguish exhaustion from truncation, it cannot report stop dates, and stop dates are what change feeds, flight timelines and duration analysis are all built from. This is the question to put to any vendor first, and the one they answer least often.

A monitoring routine that survives contact with the libraries

1. Confirm the advertiser identity on each platform separately, and store it. Re-confirm when a competitor rebrands or acquires.
2. Fix the country scope and keep it fixed. Write it down next to every number you report, because a scope change is indistinguishable from a behaviour change.
3. Collect on a schedule rather than on curiosity. Retention limits mean the value is in the series, not the snapshot, and a gap in the series is permanent.
4. Record, per run, the source, the date, and whether the run reached exhaustion or a cap. Mark ads stopped only from a complete comparable run.
5. Review changes, not catalogues. New, stopped, restarted and edited ads are the signal; the full list is just the state.
6. Keep modelled figures visibly separate from observed ones, in the same table if you like, but never in the same column.

That routine is what AdLemur automates, including the unglamorous parts: advertiser confirmation, resumable collection, exhaustion checks before a stop date, and provenance on every stored observation. It reads the same public sources as the list above, and it says which sources are off. Agents get the identical records with provenance attached over MCP, so an assistant answering a question about a competitor can tell a measurement from a model.

How can I see all of a competitor's ads?

You cannot, from public sources alone. Meta and Google publish global commercial libraries; TikTok, Microsoft, Snapchat, Pinterest and X publish European-scoped ones; LinkedIn has no programmatic access; Reddit and Amazon publish no exhaustive library. Any complete-sounding answer is stitched together from those boundaries.

Which platforms have an ad library?

Meta, Google and YouTube, TikTok, Microsoft, Snapchat, Pinterest, X and LinkedIn all publish something. They differ in geography, retention and whether programmatic access exists, which is what makes cross-platform comparison difficult.

How do I estimate a competitor's ad spend?

You do not measure it, because no platform publishes it for commercial ads. You can model it, and you should label it as modelled. Ad duration and creative volume are observations and often answer the underlying question better.

How often should I check competitor ads?

On a schedule you can keep, because retention limits mean missed windows are lost permanently. Weekly catches launches and kills; monthly catches strategy shifts and misses tests.