How to Find a Competitor's Display Ads — And What the Free Libraries Won't Show You

The free, legal way to see them

Two free, official sources cover most of what you need. The Meta Ad Library shows every ad currently running on Facebook, Instagram, Messenger and Threads — search a brand or Page name and you get the creative, the copy, the start date, and how many variants are live. The Google Ads Transparency Center does the same for search, Shopping, YouTube and, the one people forget, display: the actual banners running across the Google Display Network, filterable by region and format.

TikTok's Commercial Content Library covers the EU/EEA only, and LinkedIn runs its own ad library for B2B creative. All four work the same way: search the advertiser, browse what's live, note which ads have run longest. None of them have a bulk export button — they're built for checking one brand at a time, which is fine for three rivals and slow for three hundred.

A worked example: what the ad tells you, and what it doesn't

Here's what that search actually gets you, worked through with real numbers instead of a hypothetical brand.

Say a competitor's display banner has been live on the Google Display Network for four months. That's a real signal — a losing ad usually gets pulled within weeks. But "still running" only proves the advertiser didn't stop it. It doesn't tell you what it cost, who converted, or whether that person would have bought anyway without the ad.

Across 30 advertisers we ran through Amazon DSP in July 2026, the book delivered 78.4 million impressions at a $4.00 CPM, a blended $1.42 cost-per-click, and a 6.04x return on ad spend, with a $5.49 blended cost per acquisition across 57,137 attributed purchases — 20.1% of them new to the brand. Some of the creative behind those numbers had run for a day; some for months. Longevity told you almost nothing about which ones actually earned their spend. That's the gap between finding the ad and knowing if it worked — and no ad library, ours or a competitor's, closes it from the outside.

Why Amazon is the exception

Meta and Google publish libraries because regulation requires it. Amazon doesn't run an equivalent for DSP — its programmatic display, video and audio buy — so there's no public place to search "competitor X" and see their off-Amazon banner running on a news site or an app.

What you can see on Amazon is narrower: Sponsored Display and Sponsored Brand placements a competitor runs on Amazon's own pages, which you'll spot by browsing product listings in your category, or by being retargeted after viewing a rival's product. That's real, but it's a fraction of what a DSP campaign actually does — most of its impressions land on exchanges and apps you'll never happen to browse into, at moments you'll never happen to be shopping.

So if you're trying to find a competitor's Amazon display ads specifically, expect to see slivers, not the whole campaign. Absence of evidence isn't evidence of absence — a brand can be running a large DSP program you simply never get served.

The mistake we've made ourselves

The mistake we see most — and one we've made — is treating "this ad has run for months" as proof it converts, then copying it outright. A long run can mean the creative works. It can also mean the advertiser is running on a fixed monthly budget nobody's reviewed, or measuring last-click and never checking whether the sales it's "driving" would have happened anyway.

Last-click attribution can't tell the difference between an ad that created a sale and an ad that happened to run next to one. Neither can a screenshot from an ad library. The only way to know if a display ad is worth copying is to test it against a holdout or a matched control on your own account, not to infer it from someone else's run time.

When you search and find nothing

If you search a competitor everywhere above and find nothing, don't assume they're not advertising. Check three things first: are you searching the parent legal entity as well as the consumer-facing brand name; are you filtering to the right country, since Meta and TikTok both scope results by region; and for Amazon specifically, remember there's no library to search at all — a blank result there just means you haven't seen one, not that none exist.

If you did find ads but the analysis stalls — you've got creative and copy but no way to tell what actually worked — that's the point to stop guessing from the outside and test on your own account instead. Cloning a competitor's angle without a measurement plan just moves the guesswork onto your own budget.

Where this leaves Amazon advertisers

None of this changes whether you use an ad library, a scraper, or nothing at all: on Amazon, the honest way to know if a display ad is working is to measure it against a holdout, not to watch how long a rival's version stays live. Dr. DSP is Amazon DSP — the programmatic buy, not the courier franchise — run as a managed product by Full Circle, which has managed more than $500M in revenue across 100+ brands. Every change we make carries the evidence behind it, a measurement plan, and a rollback trigger before it runs, which is the part competitor research alone can never give you.