← All articles
Competitor Intelligence5 min read

How to research competitor catalog ads

How to research competitor catalog ads when the library shows one sample of a template: what to log, what overlays tell you, where inference stops.

A competitor's catalog ad has been following you around for a week, so you open their page in the Meta Ad Library. One entry, one sneaker, a red discount badge, running since March. By the afternoon your deck says "they're scaling this sneaker."

Nothing on that page supports the sentence. A catalog ad is a template pointed at a product feed, and the library handed you one frame of it.

Short answer: To research competitor catalog ads, treat each library entry as one template with a product sample, because Meta's Ad Library shows no spend or per-product delivery. Log the visible products, categories, overlay treatment and template start date over several weeks. Repeated products are a hypothesis about what they push, never a sales figure.

The takeaways

  • One entry can stand for a whole feed. A catalog ad with 1,400 products behind it can still preview a single sneaker, so one look tells you very little about the range.
  • Four weekly snapshots beat one deep session. A category that shows up in all 4 across several entries is worth a test. A category seen once may be sampling noise.
  • The 30-day rule weakens on catalog ads. The start date belongs to the ad and the feed can change underneath it, so a long run dates the setup.

How do you spot a competitor's catalog ad in the library?

You infer it, since the Meta Ad Library gives you no reliable catalog filter. Brandsearch's guide to competitor catalog ads makes the same point: the free library has no funnel-type filter, so catalog versus landing page gets guessed from the creative.

The tells I look for:

  • a packshot on a plain background with the product name and price in the card
  • the same frame or badge repeated across unrelated products
  • a destination URL that lands on a single product page

Two of those and I log the entry as "probable", because a designer can build all of them by hand. A multiple-versions label settles nothing either, since dynamic creative carries it too. Meta also changes this interface often, so check what your view shows today.

What does a catalog ad entry show you?

A sample, chosen in a way you can't see. Delivery decides at serve time which product each person gets, so the feed behind one ad object can run to thousands of rows while the preview shows one or a handful.

The preview can be further off than that. In a Reddit thread from May 2026, an advertiser reported that their Advantage+ ads all showed the same plain product image in the library while Ads Manager previews looked right, and a reply called it a known limitation for dynamic campaigns. One thread is thin evidence. It's still enough reason to stop treating a catalog preview as the creative people saw.

Counting rows goes wrong too. Listings already overstate ideas on ordinary accounts, as I covered in how many ads competitors run, and a single catalog row can hold more products than the rest of the page has ads.

How do you tell which products a competitor keeps pushing?

Sample on a schedule and count recurrence. Once a week for 4 weeks, open the same advertiser and log every product you can see: date, entry, product, category, price shown and destination. Four weeks is my own rule of thumb.

Then pivot by category, since single products are too noisy to read at this sample size. A category present in all 4 snapshots across several entries is a candidate for a deliberate push. Separate entries per category are the strongest thing you'll find. A shared feed doesn't split itself into a "running shoes" ad and a "winter boots" ad, so that structure suggests somebody built a product set on purpose.

What do competitor overlays and frames tell you?

Overlays are the most readable part of a catalog ad, because one template applies across every product it renders. Confect's guide to Meta's catalog templates lists what can go on one: product name, price, sale price, review scores, sticker badges and custom labels such as delivery promises.

Log one row per template: price, strike-through or percentage off, rating, delivery promise. Then line up 5 competitors. If all of them run a discount badge, a badge won't set you apart. If nobody shows a rating in a category where buyers worry about quality, that empty column is your next hypothesis.

Record a bare packshot in a preview as "unknown", because the library may be showing a default image. And a template's presence says nothing about whether it earns: even your own account can't report ROAS by overlay type.

What does a long-running catalog ad prove?

That the setup survived. The start date on a listing belongs to the ad object, and a feed can swap products, prices and images underneath it while the date stays put.

On a static ad, I treat 30+ days as a decent signal that the creative passed its owner's kill checks. On a catalog ad the same number covers a bundle: the template, the product set logic and whichever audience it runs against. If that audience is warm retargeting, the ad had an easy crowd from day one.

Eight months on one template says the advertiser isn't unhappy enough to touch it. A template replaced inside a month tells you more, because somebody made a decision you can date.

Where does the inference stop?

At sales. The library shows no spend, no ROAS and no per-product delivery, so "they push this category" is as far as the evidence goes. A product can recur because it sells, because it has margin, because the warehouse is overstocked, or because of how the feed is sorted. You can't tell those apart from outside.

Meta's EU reach figures under the Digital Services Act don't rescue this: they are reported per ad, and one ad holds the whole product set.

So test the cheap version. Put that category into its own product set in your catalog and read your own numbers, the only sales figures here you can trust.

The weekly log stays manual. Adscalr pulls the Meta, TikTok and Google Ads Transparency libraries into one dataset updated daily, and its market map places competitors on 5 messaging axes to flag the least-occupied one. It doesn't decode catalog templates or report per-product competitor performance. More on that side in competitor intelligence.

This is the thinking behind Adscalr.

See the product →