Meta Ad Library: What to Explore Beyond the Creatives
See what the Meta Ad Library shows besides creatives: run dates, placements, countries, EU transparency data, and how to read it for competitor research.
ADS Beast editorial teamPublished Updated 12 min read
This article was prepared with AI. The ADS Beast editorial team is responsible for its content.
The Meta Ad Library is a public archive of ads running across Facebook, Instagram, Messenger and Audience Network. Beyond the creative itself, each entry shows the page name, the date an ad started running, the platforms it appears on, and whether it is still active. For political ads it also publishes spend and impression ranges, and for ads delivered to the EU or the UK it shows reach and basic targeting.
Most people open the library, screenshot a few ads, and leave. That is the smallest part of what is there. The fields around the creative tell you how long a competitor has been testing an angle, where the money is going by placement and country, and which ads have survived long enough to be worth studying. This guide walks through those fields, what they mean, and where the limits are.
Short version
- Every ad entry carries a start date, page name, platform list, and active status, not just the image or video.
- Ads about social issues, elections or politics stay in the library for seven years; ads delivered to the EU or the UK stay for one year after their last impression; other commercial ads are visible only while they run.
- Spend and impression ranges appear only for political ads. Ads delivered to the EU or the UK show reach and basic targeting instead. Standard product ads show no budget data.
- A start date of 90 days or more on a commercial ad usually means the advertiser found something that works, because few advertisers keep paying for an ad that loses money.
- The library is a research surface, not a reporting tool. It will not tell you detailed targeting, conversions, or true spend for a normal brand.
What fields does the Meta Ad Library actually show?
Each ad card in the Meta Ad Library contains a fixed set of fields: the page name, the "Started running" date, the platforms the ad appears on, and whether it is currently active. The country is something you pick yourself with the country selector above the results. That is the baseline for every ad. The library does not show targeting, conversion data, or budget for most commercial ads.
Those fields are enough to answer questions the creative alone cannot. A video tells you what the advertiser wants to say. The start date tells you whether the message is still being paid for. The country selector tells you whether the offer travels across markets or only works in one. The platform list tells you whether the advertiser is buying feed, stories, or the parts of the network most people forget about.
Treat the creative as the hook and the metadata as the evidence. When you build a swipe file, capture both. A folder of screenshots with no dates is a folder of guesses.
How long do ads stay in the Meta Ad Library?
Active ads stay listed for as long as they run. Once an ad stops, its retention depends on what it was about. Ads related to social issues, elections or politics stay in the library for seven years. Ads delivered to the EU or the UK are kept for one year after their last impression, and other commercial ads are visible only while they run.
Meta lists these rules on its Ad Library help page. The difference matters when you are doing historical research. If you want to see how a brand advertised during a past season or a past product launch, the archive helps only for EU and UK delivery, and only for about a year. Political and issue advertising is the exception, and the long window is deliberate: the archive is meant to serve as a public record.
Practical consequence: build your archive now. If an ad is worth studying, save the screenshot, the start date, and the country list while the entry is live. Relying on the library to still have it later is a mistake for commercial ads.
How do you tell how long a competitor has been running an ad?
Open the ad's detail card and read the "Started running" date under the page name. An ad that has been running 90 days or more is usually a proven winner, because few advertisers keep paying for an ad that loses money. Compare start dates across a page's ads to find the long-runners.
The logic behind the 90-day marker is simple. Ad accounts churn through tests. A creative that survives a full quarter is either still profitable or still feeding a funnel the advertiser cares about. Either way, it has cleared a bar that most ads never reach.
Two caveats. First, a long run can also mean the advertiser forgot to turn the ad off, or that it is a low-budget always-on placement. Check whether the same page is running many variants of the same angle; if it is, the long run is probably deliberate. Second, a short run is not proof of failure. Seasonal offers and event-based campaigns are supposed to stop.
When a long-running ad starts to look interesting, the next question is usually when to kill your own version of it. The thresholds for that are covered in when you have enough data to switch an ad off.
Where can you see which countries an ad was shown in?
Use the country selector at the top of the library: it shows the ads a page runs in the country you pick, so checking several countries tells you where an ad is live. For ads delivered to the EU or the UK, click "See ad details" to get reach by location, broken down by age and gender.
The country view is one of the most underused parts of the library. It answers a question that the creative cannot: is this offer global, regional, or single-market? A brand running the same ad in nine countries is telling you the offer clears local economics in all nine. A brand running it in one country may be testing, or may be constrained by shipping, language, or regulation.
For ads delivered to the EU or the UK, the reach figures add a rough sense of scale. They cover those territories only. Elsewhere, you get no reach data at all.
If you are comparing placements across markets, the country view pairs well with placement analysis. The mechanics of where spend actually lands are covered in where the budget quietly goes: Audience Network, Search Partners and Pangle.
Why do some ads show no spend or impression data?
Meta publishes spend and impression ranges only for ads about social issues, elections or politics. Ads delivered to the EU or the UK get reach figures and the advertiser's age, gender and location targeting, but no spend. Standard product ads elsewhere stay private. For those, the library shows the creative, the start date and the platforms, and nothing about budget or reach.
This is a policy boundary, not a bug. The transparency rules that created the library were written around political advertising and, later, EU delivery. A shoe brand running a spring sale in the US falls outside both categories, so its budget stays in the ad account.
The usual workaround is third-party tooling such as paid trackers. The Meta Ad Library API does not fill the gap: it covers political ads and ads delivered to the EU or the UK. Be clear about what those tools can and cannot do. They can help you estimate activity, spot new creatives faster, and monitor pages at scale. They cannot see inside a private ad account, and any spend figure they show for a commercial ad is an estimate built from signals, not a reported number.
What you can and cannot get from the library
| Data point | Commercial ads (non-EU, non-UK) | Political / issue ads | Ads delivered to the EU or UK |
|---|---|---|---|
| Creative | Yes | Yes | Yes |
| Page name and start date | Yes | Yes | Yes |
| Platforms | Yes | Yes | Yes |
| Active / inactive status | Yes | Yes | Yes |
| Spend range | No | Yes | No |
| Impression range | No | Yes | No |
| Reach figures | No | Breakdown by age, gender and location | Total, and by location, age and gender |
| Payer information | No | Yes, in the disclaimer | Yes, for EU delivery |
| Retention after ad stops | None, active ads only | 7 years | 1 year after the last impression |
Read the table as a map of what questions the library can answer. If your question is "what is this brand saying," it always works. If your question is "how much is this brand spending," it only works for political ads.
How do you turn library data into a research routine?
A repeatable routine beats occasional browsing. The library rewards people who check it on a schedule and record what they find, because entries disappear and because patterns only show up over time.
- Pick a fixed list of competitor pages. Ten to twenty is enough. More than that and you will not review them properly.
- Go through each page's ads and note the oldest start dates, to surface long-runners before the new tests.
- For every ad older than 90 days, record the start date, the countries, the platforms, and a one-line note on the angle.
- Note how many ads the page is running right now. A sudden jump usually means a new test cycle or a seasonal push.
- Re-check the list on a fixed cadence, weekly or biweekly. Compare what is new against what disappeared.
- When an ad you flagged drops out of the library, treat that as a signal, not a loss. It either finished its run or got cut.
The output of this routine is not a swipe file. It is a picture of how a competitor allocates attention across markets, platforms and angles over time. That picture is what you bring into planning.
If you are running this kind of research across several client accounts, the operational side (account structure, naming, permissions) is covered in Meta Ads Manager setup for agencies and business.
What are the most common mistakes when reading the library?
The most common mistake is treating the library as a performance report. It is a transparency archive. It shows what was published and when, not how it performed.
Other frequent errors:
- Assuming a long-running ad is profitable without checking whether it is an always-on low-budget placement.
- Comparing spend across advertisers when only one of them falls under political transparency rules.
- Ignoring the country field, then wondering why a competitor's angle does not work in your market.
- Building a swipe file with no dates, which makes it impossible to tell a current winner from a retired test.
- Forgetting that most commercial ads vanish from the library once they stop, and losing the reference material you meant to keep.
One more: reading the library as a complete picture of a competitor's advertising. It is not. It shows ads that are live, plus the limited archive described above. Paused and deleted ads are mostly outside the view.
There is a related trap on the strategy side. Watching a competitor closely does not automatically mean your own positioning is clear to the market. Why a competitor gets cited in AI answers and you do not is a different problem with a different fix, covered in why an AI answer cites a competitor, not you.
What to do next
Pick three competitor pages and run the routine above once, end to end. Note the start dates, log every ad older than 90 days, and note the countries and placements. That single pass will tell you more about their current strategy than a month of casual browsing.
ADS Beast has a competitor view that shows competitors' live ads, site changes and search positions in one place: see competitor tracking.