Ad research tools, compared without the marketing

Searchable ad archives

Enormous databases you query. You arrive with a question (an angle, a niche, a product) and search millions of ads for anyone who ran something like it. Coverage and filters are what you are paying for.

Dropshipping product research

Built to find a product to sell, not a competitor to beat. Ads are the signal these tools read, but the output is a winning SKU and a supplier, and the pricing is metered in credits per lookup.

Creative analytics and ad ops

Connect your own ad accounts, tie creative to performance, and give the creative team a brief. These read your first-party data first and treat competitor ads as reference material.

“Ad spy tool” describes three unrelated jobs

Search for one of these products and you get roundup articles ranking all of them against each other on a single axis, usually database size. That ranking is close to meaningless, because the tools are not trying to do the same thing.

There are really three jobs hiding under the phrase. The first is search: you have a question and you want a very large archive to answer it. The second is product research: you are trying to find something to sell, and ads are the evidence you read to find it. The third is creative analytics: you connect your own ad accounts and learn which of your creatives actually worked.

A tool that is excellent at one is usually mediocre at the others. The price differences between them, from a free tier at one end to several hundred dollars a month at the other, mostly reflect which job they chose rather than how good they are. Work out your job first and the shortlist writes itself.

The job none of these were built for

There is a fourth question that people bring to all of these tools and none of them answer well: what changed.

Not what exists, not what is selling, not how your own ads performed. What did these five specific companies, the ones you actually compete with, do since you last looked. Which creatives went live on Tuesday. Which offer they quietly retired. Which ad has now run for four months, which is the closest thing to proof that it works.

Every tool on this page can be made to approximate that with enough manual labour: run the same searches weekly, keep the results, compare by hand. People do exactly that, with a calendar reminder and a folder of screenshots, and it holds together for about a month. The reason it collapses is that it is a scheduling problem wearing a research problem's clothes — and diffing runs automatically is what we built instead.

It is also worth saying that a lot of this data is free. Meta, Google, TikTok and LinkedIn all publish public ad libraries you can search without paying anyone. What you pay a vendor for is history the platforms throw away, filters the platforms do not offer, or somebody watching when you are not — and it is worth being clear with yourself about which of the three you are buying.

How to choose, in one question

Can you name the advertisers you care about?

If you cannot — new niche, new market, no idea yet who the players are — you need discovery, and a monitoring tool that demands a name up front is useless to you. Go to a searchable archive, or to a product-research tool if what you are hunting is a SKU rather than a rival.

If you can name them, discovery is the wrong purchase. You do not need a bigger haystack; you need those specific brands watched on a schedule, their creatives archived before the libraries drop them, and a message when something moves.

Most teams eventually want both, and they are not really substitutes — one is a research subscription you dip into, the other is a standing service that reports to you. But if you are buying one thing this quarter, that question sorts it.