How to use an ad library for product research
Public ad libraries show what stores are paying to promote right now. Here is a repeatable method for turning Meta, TikTok, Pinterest and Snapchat ads into a shortlist of products worth testing.
On this page
Every store that pays for ads leaves a public trail. The major ad platforms publish libraries of the ads running on their networks, and those libraries are the closest thing ecommerce has to a live list of what other sellers are betting money on. This guide shows how to search them on purpose, how to tell a test from a product that is working, and how to turn what you find into a short list you can validate.
Key takeaways
- An ad library tells you what a store is promoting, not what it is selling. Treat every ad as a lead to check, not proof.
- Search by problem and by product type, not only by product name. Sellers rarely use the words you expect.
- Look for repetition. Several ads for one product, several stores running the same product and ads that keep running are stronger signals than one polished video.
- Always follow the ad to the store. The landing page, price and catalogue tell you more than the creative does.
- Write down what you find in the same format every time, so you can compare candidates fairly.
What each library shows
Meta Ad Library
The Meta Ad Library covers ads on Facebook, Instagram, Messenger and Meta's partner network. You can search by keyword or by advertiser page, filter by country, and see when each ad started running. Ads that share a creative are often grouped as versions, which tells you the advertiser is testing copy or formats around one idea. For ads shown in the European Union, Meta also publishes extra transparency detail, including an EU reach figure and a breakdown by age and gender. Outside the EU, most commercial ads show far less.
TikTok Creative Center
The TikTok Creative Center is built for advertisers looking for inspiration. Its top ads section lets you filter by industry, region and objective, and its trend tools show which hashtags, songs and topics are rising. It is a good place to learn which formats and hooks are working in a category, but it is a curated selection rather than a full list of every ad. TikTok also runs a separate commercial content library for ads shown in Europe.
Pinterest and Snapchat
Both Pinterest and Snapchat publish ad transparency libraries, with the most detail available for ads shown in the EU. They hold fewer ecommerce advertisers than Meta, which is useful in its own way. A product that is advertised on Pinterest usually has a visual, aspirational angle, and a product pushed on Snapchat is usually aimed at a younger audience.
RivalSonar ad library
The RivalSonar ad library collects public ads from Meta, TikTok, Pinterest and Snapchat in one search. Each ad shows the advertiser, the store it links to, how many days it has been active, an estimated reach and an estimated spend, plus an AI score that labels the ad as testing, scaling or winning. Because ads are linked to stores, you can jump from an ad to the store profile and see its estimated sales, traffic and best sellers without leaving the page.
A step-by-step method
Step 1. Pick a lane before you search
Open-ended browsing feels productive and rarely is. Decide on one niche or one customer problem for the session, such as "back pain at a desk", "dog hair on furniture" or "small kitchen storage". Write it at the top of your notes.
Step 2. Search with problem words, not product words
Sellers write ads around the pain, not the product. Try searches like "finally", "tired of", "no more", "best purchase" and the problem itself. Then search the product type in plain words. Run the same searches in two or three countries, because a product can be saturated in one market and barely tested in another.
Step 3. Filter for ads that have survived
Fresh ads are mostly tests. Sort or filter for ads that have been running for a few weeks. A store that keeps paying for an ad over weeks is getting something back from it. In RivalSonar, filter by active days and by the AI score, and start with ads marked scaling or winning. The guide on reading ad longevity and reach explains how to judge this signal without over-trusting it.
Step 4. Count repetition
Open the advertiser and count how many separate ads point to the same product. One ad is a guess. Five to ten variations with different hooks usually means the product is selling well enough to justify more creative work. Then check whether other stores advertise the same product. Several stores competing on one product confirms demand, though it also tells you the market may be getting crowded.
Step 5. Follow the ad to the store
Click through and look at:
- the price and any compare-at price
- whether there are quantity breaks or bundles
- how many products the store sells and whether it is a one-product store
- shipping times and the returns promise
- the reviews on the product page, and whether they look real
In RivalSonar, the store profile adds estimated monthly visits, estimated orders and revenue, the store's best sellers and the other ads it runs. Remember these are model estimates, so use them to rank candidates, not to forecast your own results.
Step 6. Record each candidate the same way
Keep a simple table so you can compare products side by side.
| Field | What to note |
|---|---|
| Product | Name and a link to the best example store |
| Problem it solves | One plain sentence |
| Ad evidence | Number of ads, longest active run, networks |
| Stores selling it | How many you found, and in which countries |
| Price range | Lowest and highest you saw |
| Store estimates | Estimated monthly orders for the strongest store |
| Notes | Angles, weak spots in competitors' pages |
Saving the ads and stores to a collection in RivalSonar does the same job and keeps the links in one place. A saved search with a daily digest will email you when new ads match your query.
What ads cannot tell you
Ad libraries show spend, not profit. A store can run an ad for a long time and lose money on it, especially if it is chasing customer lifetime value or building an email list. Large brands also keep ads running for brand reasons that have nothing to do with one product. And libraries only cover paid social, so a product that sells mostly through search or influencers may barely appear.
That is why ad research should feed into validation, not replace it. Once you have three to five candidates, work through the product validation checklist and, for the strongest ones, look at the store's real catalogue movement with sales tracking.
How credits apply
In RivalSonar, a search costs one credit per result row and opening an ad or store costs two credits, after which that page is free to re-open for 24 hours. Narrow your filters before you search wide, and you will spend far fewer credits. Every plan, including the free plan, comes with monthly credits. The pricing page has the current allowances.
