Meta publishes every ad that is currently running on Facebook, Instagram, Messenger and its Audience Network in a free, searchable archive: the Meta Ad Library. No account is needed to search it. For product research it answers one question better than anything else: is someone paying, week after week, to advertise this product?
What the library shows, and what it does not
| It shows | It does not show |
|---|---|
| Every active ad of an advertiser, with its picture or video and text | How much was spent on an ordinary commercial ad |
| The date each ad started running | Clicks, sales or ROAS |
| The platforms it runs on | Who it is targeted at, outside the EU |
| How many versions of the ad exist | Ads that have been switched off, unless you filter for them |
| Where the ad sends people (the link on the ad) | Which of the versions is the one that works |
So the library gives you behaviour, not results. That is enough, because advertisers behave differently when an ad pays.
The three signals
How long the ad has been running
Each ad shows the date it started. Somebody pays for every day an ad stays on. A small store that has kept one ad live for six weeks is almost certainly earning from it, because a small store cannot afford not to. Ads a few days old tell you nothing yet: every test starts that way.
How many versions exist
The library groups near-identical ads and says how many versions there are. Many versions of one idea means the advertiser is spending on finding the best cut of something that already works. One version of twenty different ideas means they are still searching.
How many ads the advertiser runs for the product
Open the advertiser’s page in the library and count the active ads that point to the same product. A store with thirty active ads for one product has made it the centre of its business. A store with one ad each for thirty products is testing, and you should read its ads as experiments, not as proof.
A 15-minute routine
Search the product, not the brand
Choose the country you would sell in, set the category to all ads, and search the plain words a buyer would use: “posture corrector”, not a brand name. Keyword search looks at the text of the ads.
Filter to active ads and sort your attention by start date
Scroll for the oldest start dates among active ads. Those are your evidence. Note the advertisers behind them.
Open each of those advertisers
Look at how many active ads they run and for how many products. Follow the link on the ad to the store and note the price and the offer.
Write down what the long-running ads have in common
The first three seconds, the claim, the offer, the format. If the survivors all open on the same demonstration, that demonstration is what sells the product.
Count the newcomers
Go back to the search and count advertisers whose ads all started in the last two weeks. A few is normal. A wave of them with the same clip is the crowding signal described in six checks for saturation.
Mistakes to avoid
- Reading a big brand like a small store. A large company can run an unprofitable ad for months for reasons that have nothing to do with you. The longevity signal is strongest for small advertisers.
- Copying the ad. The ad that has run for two months has also been seen for two months. Take the idea that works, the problem shown and the result, and make your own version of it.
- Stopping at Facebook. A product can be quiet on Meta and busy on TikTok or Pinterest. Check the others before you call it undiscovered.
- Treating one day as the picture. The library shows what is live at the moment you look. Check again a week later: the ads still running are the ones that matter.
Where a tool saves time
Everything above can be done by hand, and it is worth doing by hand a few times to learn what the signals look like. The slow parts are searching three ad platforms separately, keeping track of how long each ad has run, and connecting an ad to the store and the product behind it. NicheHunter keeps ads from Facebook, TikTok and Pinterest in one search with the days each has been running, and links each ad to its store and, on TikTok Shop, to the product’s units sold. You can look around free, and the method for choosing what to search for is in how to find a winning product.



