Competitor ads, tied to what actually sold
Ad libraries show you a rival's creatives. Type a catalog store — if it is not watched yet we still probe and log it. For stores with a daily watch, every Meta, Google, and TikTok ad sits next to unit estimates.
Try it
Live search
Runs on this page. You do not need a key.
Looking that up…
A worked example: MagBak
MagBak sells phone cases and mounts. Over 51 observed days in mid-2026, the watch recorded an estimated $1.24 million in revenue across 20,084 units, with a peak day of 450 units on July 5. On the ads side: 729 Meta creatives on record with 30 live, plus 410 on Google with 23 live. The interesting part is the join — on days with heavy ad activity, the iPhone 15 case sold 82.7% more units than on quiet days. The iPhone 16 version lifted 22.9%, the iPhone 14 version 34.9%. And when new Meta creatives appeared on August 20 pushing Pixel 11 cases, that showed up in the feed the same week — a product-line bet, announced by the ads themselves.
None of that requires guessing what a competitor spends. Run length, creative count, and the sales curve next to it say more than a spend estimate ever could.
What an ad watch returns
Put a Signals watch on a store and its ad picture assembles daily: every creative found on Meta, Google, and TikTok, when each was first and last seen, how long it has run, and which product it points at. On top of the raw library you get the correlation views — ad activity against daily units, revenue attributed per creative where the ad links straight to a product, and per-SKU lift on high-ad versus low-ad days. Watches are part of paid plans; the Explorer plan starts at $29 a month with 8 watch slots.
From their winning ad to your brief
The point of watching competitor ads is not admiration. Connect the catalog to Cursor or Claude and the ad data becomes working material:
“Which of magbak.com's ads has run the longest, and which product does it push?”
Run length is the closest public signal to "this ad works" — nobody pays to keep a losing creative alive for months.
“Their new campaign started last week. Did units move?”
The ads timeline lines up creative launches against daily unit estimates, so "did it work" is a lookup, not a hunch.
“Take their three longest-running ads and draft versions for my store, my products, my voice.”
This is the copy-the-competitor move done right: borrow the angle that demonstrably runs, not the words.
“Which of their products gets ad support but isn't selling?”
Ad-supported SKUs with flat units are money bleeding — theirs to waste, yours to learn from.
What you can check before paying
The live lookup above probes the store you type. No watch yet? The request is logged and the ads crawl is queued. Daily unit estimates of your own are paid. The 14-day trial still shows plenty about a competitor for free: the store card carries a monthly ad-spend estimate and a traffic split by channel, the change feed shows their price moves and new products, and the price feed catches their markdowns. If a store advertises hard and reprices often, you will see it there first.
Things people ask
Is this an ad spy tool?
It shows competitor ads, so partly yes — but ad libraries already do that for free. The difference is the other half: each watched store's creatives sit next to daily unit estimates, so you see which ads ran alongside actual sales movement, not just which ads exist.
Which ad networks are covered?
Meta, Google, and TikTok creatives for watched stores: format, first and last seen dates, run length, and which product the ad points at when the link carries a product URL.
Can my AI write ads from this?
That is the workflow this data is for. Connect Neonjelly to Cursor or Claude, pull a rival's longest-running creatives, and ask the agent to draft your version — same angle, your product and voice. The agent works from real creatives and real run lengths, not memory.
Is ad tracking on the free trial?
Type any catalog store. If it is not on a daily watch yet, we still queue an ads probe and log the request — creatives fill in as the crawl lands. Daily unit estimates stay on paid watches, from $29 a month. The trial still covers store cards with an ad-spend estimate, plus price and product tracking across 1.37 million stores.
