The Meta Ad Library is a public archive of the ads currently running on Facebook and Instagram. No account required, no fee, and no need for your competitor to know you are looking. We show how to narrow a search to Instagram only in our competitor Instagram ads article.
Most people still open it once, type a brand name into the search box, see a messy list and close the tab. The problem is not the tool. It is which question you brought to it.
Pick the right tab first
The library has two modes and they show completely different things:
- All ads — commercial advertising. This is almost certainly what you want. It shows active ads only; a paused campaign does not appear here.
- Issues, elections or politics — political and social-issue advertising. The archive is far deeper here: spend, reach and demographic data are included, and stopped ads remain visible.
If you are doing competitor research you are in All ads, and you have to accept one thing: everything you see is live today. The library is a snapshot, not a history.
That single sentence is what most people miss: if an ad you saw yesterday is gone today, it was not deleted — it was turned off. And that fact alone is worth something.
The country setting decides your results before you search
You choose a country before you search. Two mistakes are very common here:
- Leaving the default alone. The library picks a country from your location. You intend to study the German market and end up reading Turkish results.
- Looking at one country and concluding "this brand runs 4 ads". The same brand runs entirely different creative in different countries. Change the country and the list changes.
The country filter narrows where the ad is shown, not where the advertiser is based.
Brand search and keyword search are not the same thing
There are two routes through the search box, and confusing them is the most expensive mistake:
| Search type | What it finds | When to use it |
|---|---|---|
| Advertiser | Every active ad from one specific page | Tracking a competitor you already know |
| Keyword | Every ad whose copy contains that word | Finding out who advertises in a category |
Keyword search is the only way to find competitors whose names you have never heard. Search for offer phrases — "free shipping", "30-day trial", "limited stock" — not brand names. You learn how a whole category talks in a single session.
The three fields worth reading
Every card carries three genuinely useful pieces of information, and all three are easy to skim past:
- Launch date — "Started running on 1 Jan 2026". How long has this ad been going? It is the strongest indirect signal of performance you get.
- This ad has multiple versions — the line saying the same copy is running with 12 different images. A high number means active testing.
- Call to action —
Shop Now,Learn More,Sign Up. It tells you which stage of the funnel the ad targets.
An ad that has been running a long time, has many versions and carries Shop Now is an ad your competitor believes is making money. Nobody runs a losing creative for months.
What the library will not tell you
To be honest about it, the library has real limits:
- No spend. Commercial ads show no budget, reach or impression data. That exists for political ads only.
- No performance. Click-through, conversion, whether it worked at all — none of it.
- No audience. You cannot see who was targeted.
- No history. A paused ad drops off the list.
Which is why serious competitor research is a repeated measurement, not a single look: record the same competitor twice, two weeks apart, and the difference between the two is the performance signal.
A routine that works
- Write down your competitor list — 5 to 10 pages. More than that cannot be tracked.
- Set the country to your target market for each one.
- Scroll the whole list and save it. Reading on screen is not enough; you will compare it in two weeks.
- Repeat two weeks later. Look at what disappeared and what appeared.
- Collect the copy of the long-running ads somewhere separate. That is where the category's winning patterns live.
Step four is the entire job. You learn nothing on the first pass and everything on the second.
The three most common mistakes
- Looking only at the image. The most copyable part of an ad is its first two lines, not its picture.
- Focusing on a single competitor. What one brand does may be a preference; what five brands do at once is a pattern.
- Stopping the scroll halfway. The library uses infinite scroll, and the order on the first screen is not a performance ranking. If you do not reach the end, your sample is random.
The third is the sneakiest: drawing conclusions from the first 20 ads of a 200-ad list is like ending a survey after 20 respondents.
To do this with the extension: Meta Ad Library guide · Meta Ad Library auto scroll