Meta Catalog Ads: Catalog Setup, Feed Errors and Match Rate

What are Meta catalog ads and how do they work?

Meta catalog ads, officially called Advantage+ catalog ads and still widely known as dynamic product ads (DPA), pull product images, names and prices straight from your catalog and decide which product each person sees. You don't design an ad per product; Meta picks the item using the signals your Meta Pixel and Conversions API (CAPI) send about browsing and buying.

That means most of the performance is decided before you ever open Ads Manager. Three pieces have to work together:

• The catalog: a complete product list that follows Meta's field rules.
• Event data: ViewContent, AddToCart and Purchase events whose product IDs match the IDs in the catalog.
• Product sets: the groups that decide which products can appear in which ad.

The same catalog powers ads on Facebook, Instagram and the rest of Meta's placements, and it can also feed a Facebook or Instagram shop where your account is eligible.

How do you set up a Facebook catalog in Commerce Manager?

You create the catalog in Commerce Manager, choose the e-commerce (online products) type, add products through a data source and connect your pixel or dataset. Menu labels shift from time to time, so the names below reflect the interface as of September 2026.

1. Create the catalog In Commerce Manager, add a new catalog, pick the e-commerce type and assign it to the right business portfolio.
2. Add a data source Go to Catalog > Data sources and add a data file (URL, Google Sheets or upload), a partner platform or the pixel.
3. Read the first upload Open the feed overview, check errors and warnings, and click Download issue report if anything failed.
4. Connect your pixel Under Catalog > Events, connect the pixel or dataset; the match rate can take up to 48 hours to appear.
5. Build product sets Under Catalog > Sets, create sets by category, price or margin before you build the campaign.

With the prerequisites in place, the campaign itself is short. In Ads Manager click + Create, choose the Sales objective, turn on Advantage+ catalog ads at campaign level and select the catalog. At ad set level, pick the product set under Promoted products, then choose single image or video, carousel or collection as the format. If collection doesn't appear in ad setup, Meta says you can still select it in the ad creative section under Format display options. Also note that, starting March 2026, single media and Advantage+ catalog collection ads in Facebook Feed no longer show a URL in the ad footer, so your product image and headline have to carry the click on their own.

Which data source should you use for your product feed?

For any store where prices and stock change often, use a scheduled feed or a partner platform integration; a one-time file upload only suits tests or very small catalogs.

Supported formats are CSV, TSV, XLSX, XML (RSS/Atom) and Google Sheets. Clean titles and attributes matter here as much as they do on Google Shopping, so the checklist in our shopping feed optimization guide is worth running before you connect the file to Meta.

Scheduled feeds come in two modes. Replace rewrites the catalog from the file, deleting products you removed, but processes more slowly. Update is fast but ignores products you simply delete from the file. Meta's own suggestion is to combine them: an hourly update schedule for prices and availability, plus a weekly or monthly replace schedule to clear out old products. The two schedules for the same feed must run at least 10 minutes apart.

How do you fix Facebook product feed errors?

Most feed errors come from a missing required field, a badly formatted price or Meta being unable to reach your file. Go to Catalog > Data sources, click the feed and read the summary of the last upload session, then download the issue report for a sample of each problem.

Start with the required fields. According to Meta's catalog field reference, every e-commerce item needs id (up to 100 characters, unique), title (up to 200 characters, 65 recommended so it isn't cut off), description (up to 9,999 characters, plain text), availability (in stock or out of stock), condition (new, refurbished or used), price, link, image_link and an identifier; Meta's troubleshooting guide asks for at least one of brand, gtin or mpn. Images must be JPEG or PNG, at least 500 × 500 pixels and no larger than 8 MB.

If your platform exports prices with thousands separators or the wrong field names, fix it at the source when you can. Otherwise add data rules in the feed's settings to change capitalization, map a column such as "sku" to id, find and replace values or set defaults.

What is catalog match rate and how do you improve it?

Catalog match rate is the share of product IDs in your pixel or app events that match a product in the connected catalog. It depends on the content_ids (or contents) your ViewContent, AddToCart and Purchase events send being exactly the same as the id values, or group IDs, in your catalog.

You'll find it in Commerce Manager under Catalog > Events. Meta calculates it over the last 28 days of data, with up to a 48-hour delay.

Worked example: over the last 28 days your pixel sent 2,000 content IDs and 1,500 of them exist in your catalog. Your match rate is 1,500 ÷ 2,000 = 75%, the bottom of Meta's recommended range, and one in four product IDs your events sent can't be used for targeting. When you download the issues report for the other 500 IDs, the causes are usually one of these:

• Variant vs parent mismatch: the pixel sends an ID that exists in the catalog as neither an id nor an item_group_id, for example your platform's internal product number while the catalog uses variant SKUs. If you send group IDs, set content_type to product_group.
• Formatting differences: "SKU-1042" in the catalog, "sku-1042" in the pixel; IDs are case sensitive. Also check the parameter is spelled content_ids and that multiple products are sent as ['12345','67890'], not '12345,67890'.
• Products missing from the catalog: items that never made it into the feed, were archived or were rejected for errors.

Browser pixels miss events because of ad blockers and cookie restrictions, so send the same events server-side through CAPI and deduplicate them with a shared event_id. Cleaner data also helps your ad sets get through the learning phase faster.

How do you create product sets for catalog ads?

A product set is a subset of your catalog used in a specific ad. Create one in Commerce Manager under Catalog > Sets > Create set, using filters or by picking products manually. Every catalog also has a default "All products" set that can't be edited.

Filter-based sets are dynamic. Rules like "product type contains women's shoes" and "price is less than 100 USD" pull in new matching products automatically and drop ones that no longer qualify. Manual sets stay frozen until you edit them. For catalog ads, filters are almost always the better choice.

Product sets that earn their keep

• Margin sets: write a label such as "high-margin" into custom_label_0 and filter on it. It's the simplest way to steer spend toward profitable products.
• Category sets: splitting categories with different basket sizes lets you read ROAS per category.
• Cross-sell sets: one set is shown in the ad, a second defines the audience, for example ski boots for people who recently bought skis.

Don't slice too thin, though. When you build a set in Commerce Manager, a product selection indicator in the bottom right shows its size, and Meta says broader sets can give delivery more room to find the right product. The "Add recommended products" toggle on the same screen adds bestsellers and products with conversions in the last 14 days even if they don't match your filters. Avoid editing a set that a live campaign is using, since the audience update can disrupt performance; use Create similar set and launch it in a new ad set instead.

Should you use catalog retargeting or broad targeting?

Retargeting usually pays back fastest if your site gets steady product views and add-to-carts each month, while broad targeting with the same catalog is how you find new customers. For most stores the right answer is both, in separate ad sets with separate budgets.

Retargeting runs on catalog custom audiences. In the ad set's Audience section, click Show settings, then include a catalog custom audience based on a product set and an event (viewed, added to cart or purchased). Recent purchasers go under Controls > Show more controls > Exclude these custom audiences. For prospecting, keep targeting broad rather than stacking interests.

Example split: a $100 daily catalog budget could run as three ad sets: added to cart 0–3 days ago, viewed a product 4–14 days ago, and broad. Start with the smaller share on the warm segments and the larger share on broad, then compare cost per purchase by segment after two weeks. Treat that split as a hypothesis, not a rule: if the warm audiences are tiny, their budget won't spend and frequency climbs fast.

When and how to judge the results

Give catalog ads at least two weeks and let them exit the learning phase before deciding anything, and judge them on cost per purchase against a margin-based ROAS target (see what ROAS is and how to calculate it), not on cost per click. Check that the most-shown products are the ones that actually sell and that no spend goes to out-of-stock items. Platform attribution can also credit catalog ads, especially retargeting, with sales that might have happened anyway, so compare Meta-reported purchases with your store's orders. Marpany's ad performance view puts Meta, Google and store data in one table so that gap is visible.

Your next steps

Whether you're launching or auditing, work through this list in order.

• Confirm all products live in one catalog and that it's connected to the right pixel or dataset.
• Open Data sources, check error and warning counts on the last upload and download the issue report.
• Verify prices look like "34.90 USD" and every product has brand, GTIN or MPN.
• Set an hourly update schedule for price and stock, plus a weekly replace schedule for clean-up.
• Read the match rate under Catalog > Events; below 75%, fix your content_ids format first.
• Run CAPI alongside the pixel with event_id deduplication.
• Create filter-based sets: at least one category set and one high-margin set.
• Launch retargeting and broad in separate ad sets and compare cost per purchase after two weeks.