The free method, step by step
- 01Pick 10 to 20 advertisers in your niche. Competitors, adjacent offers, and anyone whose funnel you respect.
- 02In the Meta Ad Library, list their active ads and read the start dates.
- 03Flag every ad running more than 30 days. That is your shortlist · the ads that survived their own testing.
- 04Look for the same hook across multiple creatives. A repeated angle is a validated angle.
- 05Separate the hook from the format. « Show the dashboard in the first two seconds » is a format; « you are guessing and it is costing you » is the hook. Hooks travel between niches, formats do not.
- 06Read the first three seconds and the first line of copy. That is where the decision is made, and it is the part worth studying.
- 07Write down the offer each ad points to. The same hook in front of a different offer is a different experiment.
Duration beats everything else as a free signal. Ignore like counts · they measure nothing about profitability.
Where this stops working
- You have to compile the advertiser list yourself, and you will miss the ones you have never heard of · which are the interesting ones.
- Media expires. The ads that built the business are exactly the ones that no longer load.
- No history, so you cannot see who is scaling right now. Scaling is the strongest buy signal there is, and it is invisible from a single visit.
- No way to save, tag or compare. The swipe file lives in your screenshots folder.
- By the time you have done this for a niche, the top of it has changed.
How LaunchSpy does the same thing
- Search ads by niche, country, platform and language rather than by advertiser name
- Ad media kept, so an old winner stays studyable
- Duration and history per ad, and per creator
- Who started scaling in the last 30 days, surfaced rather than hunted
- Boards and swipe files, so the research survives the session
Questions people ask
How long does an ad have to run to count as a winner?
Thirty days is a reasonable line for info products. Under a week it is a test. Between one and four weeks it is undecided.
Can I just copy a winning ad?
Copy the structure, not the asset. The hook, the sequence and the offer logic are what transfer. The exact copy and creative are someone else's property, and lifting them is both a legal problem and a bad ad.
Do likes and comments tell me anything?
Almost nothing about profit. Plenty of high-engagement ads lose money, and plenty of quiet ones print. Duration is the honest signal.
Do this for a whole niche, not one page at a time
110,000+ creators · 15.2M+ ads · 260,000+ funnels · 51 platforms · 10 markets
Start free