How to read a competitor's Meta Ad Library

What the Ad Library shows, and what it doesn't

Meta's own description:

“People can search for all active ads running across products from Meta.”Meta Transparency Center, Ad Library tools

• Every active ad, from any advertiser, across Facebook, Instagram and Meta's other apps.
• More detail for some ads. Ads about social issues, elections or politics stay in the library for seven years, with who paid, a spend range and reach. Ads delivered in the EU also carry extra information and are kept for a year after they stop.
• Not shown for ordinary ads: spend, results or targeting. You see what a competitor is saying, not how well it's working.

Sources: Meta Help Centre, Meta Transparency Center.

How to read it, step by step

1. Find the right Page. Search the brand, then choose the advertiser's Page rather than a keyword search. Big brands often advertise from several Pages (regional, product or parent), so check each one.
2. Stick to live ads. They show what the competitor is paying to run today.
3. Note how long each ad has been running. Every ad carries the date it started. An ad that has run for months is usually one the advertiser keeps choosing to fund. That's a reading, not proof of results.
4. Collapse the duplicates. The same ad often appears many times, re-posted or re-cropped. Count it once.
5. Group what's left into ideas. Ask of each ad: what does it claim, and who is it for? Ads that make the same claim to the same buyer are one idea, whatever they look like. The briefing guide has the rules we use.
6. Look at the spread. How many ideas are there, and how much of the library does the biggest one carry? That tells you what the competitor is betting on, and where they're thin.
7. Put yours next to it. The useful question isn't “what are they running?” but “what are they saying that we aren't, and the other way round?”

Where reading it by hand breaks down

• Volume. Large advertisers run hundreds or thousands of live ads. Grouping them by hand takes days.
• Order. The library doesn't list ads newest first. In our tests, the first hundred ads of a big library gave a misleading picture: one bank's first 120 ads showed 12% launched in the last 30 days, against 57% across its first 620.
• Judgement. Deciding whether two ads are one idea or two is a call, and one person's grouping drifts. We group every account three separate times and report the middle count.