Meta Ad Library is a free, public archive of the ads running across Facebook, Instagram and Meta's other apps. It shows the creative, the copy, the platforms and the date each ad started. It does not show spend, results or targeting for ordinary commercial ads outside the EU, so it tells you what a competitor is running, never what is working.
That one sentence is the reason most competitor research built on Ad Library goes wrong. The rest of this page is the detail: what the library shows, the nine things it cannot show, where the EU is the exception, and a research workflow that stays inside what the data can support. The competitor ad analysis hub walks through the same workflow from search to original creative.
What Meta Ad Library shows
| Data point | Commercial ads, outside the EU | Commercial ads delivered in the EU | Social issue, election or political ads |
|---|---|---|---|
| Creative and ad copy | Yes | Yes | Yes |
| Start date and platforms | Yes | Yes | Yes |
| Active status | Yes | Yes | Yes |
| Inactive ads | Not documented by Meta; practitioners report they drop out | Archived for 1 year after the last impression | Stored for 7 years |
| Targeting parameters chosen (age, gender, location) | No | Yes | Via the Ad Targeting dataset for researchers |
| Who was reached (demographic and regional breakdown) | No | Yes | Yes |
| Spend | No | No | Yes, as a range |
| Clicks, conversions, ROAS | No | No | No |
| Available through the Ad Library API | No | Yes | Yes |
Sources: Meta Ad Library tools, Transparency Center and Meta's 2023 DSA announcement.
Two rows in that table get misread most often.
Spend is only published for political and social issue ads. EU transparency rules added targeting and reach data for commercial ads delivered in the EU. They did not add spend. A spend figure for a commercial advertiser did not come from Meta.
The API does not cover most commercial advertising. Meta describes the Ad Library API as covering ads about social issues, elections or politics, and ads delivered to the EU and associated territories. A brand advertising only in the United States can be browsed on the Ad Library website but cannot be pulled through the API.
One 2025 change also matters if you research EU markets. Meta stopped allowing political, electoral and social issue ads in the EU from October 2025, in response to the EU's political advertising regulation. The richest data category in the library no longer exists for new EU ads.
The 9 things Ad Library cannot tell you
Each of these gaps changes how a competitor's ads should be read.
- 1.Spend. No figure and no range for ordinary commercial ads, in any region.
- 2.Performance. No impressions, click-through rate, conversions or return on ad spend. An ad that has run for eight months might be a proven winner or an ad nobody remembered to switch off.
- 3.Targeting, outside the EU. For ads that never reached the EU, who the ad was aimed at is not shown. A targeting description built from Ad Library alone is a guess.
- 4.Budget concentration. You can see ten ads. You cannot see that one of them takes most of the budget. The visible set looks like a portfolio when it may be one or two ads that matter and a tail that barely spends.
- 5.Campaign structure. Objective, budget strategy, ad set layout and bid strategy are invisible. Two advertisers running near-identical creative can be set up in opposite ways.
- 6.Everything after the click. Landing page, offer mechanics, checkout, email follow-up and price tests sit outside the frame. The ad is the smallest part of the system it belongs to.
- 7.Test design. Five similar ads might be a deliberate test or five unrelated assets from different freelancers. You cannot tell which.
- 8.Why an ad stopped. A disappeared ad may have been paused for poor results, ended with a promotion, hit a policy issue, or been replaced by a new version. The library records that it stopped, not why.
- 9.Complete history. Outside the EU and the political category, Meta does not document how long inactive commercial ads stay visible, and gaps are not marked as gaps. What you can see today is not what you could have seen six months ago.
Why a long-running ad is not proof of profit
The most repeated advice about Ad Library is to sort competitor ads by how long they have run and treat the old ones as winners. An earlier version of this article gave the same advice. We have withdrawn it, because the reasoning does not hold up.
Run time has several causes that look identical from outside. Steady profitable delivery, an account nobody is watching, an evergreen brand ad on a small budget, a retargeting ad that barely spends, and a contract with a creator all produce the same long start date.
Spend is concentrated and you cannot see where. When a few ads absorb most of an account's budget, any single ad you pick is, by base rates, more likely to be in the low-spend tail.
Their economics are not yours. An ad that pays back at their margin, order value and repeat rate can lose money at yours. Copying the creative imports none of the conditions that made it work.
What run time does support is a smaller claim that is still useful: the advertiser has chosen to keep this message in market. That is evidence about their positioning. Treat a long-running ad as a lead worth inspecting, then test the idea behind it in your own account.
What Ad Library is genuinely good for
None of these needs spend or performance data.
- 1.Claim mapping. Which promises does the category make, and which promise is nobody making? This is the most valuable thing the library does.
- 2.Format census. Static versus video, polished versus native, aspect ratios, caption styles. All observable and reliable.
- 3.Offer intelligence. Discount depth, bundles, guarantee wording and when each appears in the calendar.
- 4.New entrants. Who started advertising in your category recently.
- 5.Exhausted angles. Seeing the angles every competitor already uses tells you where the open ground is.
- 6.Compliance reference. How others in a regulated category word claims that passed review.
What a paid tool can add, and what nobody can
| Gap | Can a third-party tool close it? |
|---|---|
| History beyond what Meta keeps visible | Yes, if the tool recorded the ad while it was live |
| Search, filters and tagging at scale | Yes |
| Alerts when an advertiser launches or stops ads | Yes |
| Patterns across many advertisers | Yes |
| Competitor spend for commercial ads | No. Any figure is an estimate |
| Competitor results | No. Not published |
| Competitor targeting outside the EU | No. Not published |
The top half of the table is real work worth paying for. The bottom half is not available to anyone, because Meta does not publish it. When a tool shows competitor spend or performance, ask how it was estimated. A labelled estimate with a stated method can be useful. An estimate presented as observed data is a different thing.
AdRiseLab competitor intelligence works inside the same public boundary. It reads publicly visible Ad Library data, groups hooks, formats and run time, and does not have access to competitor spend, results or targeting.
A 6-step research workflow that respects the limits
Work through the steps in order; each one narrows what the next has to look at.
- 1.Census the category. List every advertiser in your competitive set and every ad currently running. Record format, claim, offer and start date.
- 2.Map claims, not ads. Group ads by the promise they make. The output is a map of crowded and empty positions.
- 3.Record run time as commitment, not results. A long start date says the advertiser kept the message in market, nothing more.
- 4.Track offer timing across the calendar. Offer structure and its timing are among the most reliable signals in the library.
- 5.Turn observations into hypotheses and test them in your own account. This is the only step that produces evidence about your economics.
- 6.Never copy an execution. Beyond the trademark and originality risk, the execution is the part most dependent on their brand and their audience's history with it.
For step 1, the free competitor ad analysis worksheet has 17 columns, including one for what you cannot know. For step 5, write each test with the free creative brief template so the result is readable. The competitor ad analysis hub covers the workflow in more depth.
Sources
Meta Transparency Center: Ad Library tools. Meta Ad Library. Meta newsroom: new transparency measures as the Digital Services Act comes into effect, August 2023. Meta newsroom: ending political, electoral and social issue advertising in the EU, July 2025. Meta advertising standards.
Related Reading
For how the ads you launch from this research should be structured, see the static ad anatomy with 9 layout templates and 12 Meta video ad script templates. For whether a competitor idea actually beats yours, use the creative testing budget calculator. AdRiseLab plans start at $39 a month; see pricing.