Product feeds for Google Shopping and AI answers: a checklist.

Why the feed matters now.

Google’s shopping results, free listings and AI shopping answers draw on product data from Merchant Center. There is no separate feed for AI.

Google’s product structured data guidance says that providing both structured data on your pages and a Merchant Center feed gives you the most eligibility for shopping experiences. Thin titles and missing attributes make it harder for Google to match a product to the specific question someone asks.

The attributes every product needs.

Brand is required for new products, and a correct GTIN is strongly recommended where one exists. Google warns that products without one may have limited visibility, and that an incorrect GTIN can get a product disapproved.

What has to match.

Variants and apparel.

• Group variants with the same item group ID
• Give clothing and shoes a size
• Give apparel and accessories a color, age group and gender
• Flag bundles and multipacks where they apply

Titles and descriptions that answer questions.

A title like Rain Shell tells Google little. A title that names the product type, the key attribute and the use, such as Two-layer rain shell, packable hood, coastal hiking, gives Google more to match against. Descriptions should say what the product is and who it is for in plain words, without claims your product data cannot back up.

Keeping it healthy.

1. Read diagnostics every cycle. Fix disapprovals at the source, in the store, so the fix holds.
2. Check a sample against the store. Price, stock and link for a handful of products, every time.
3. Improve titles in batches. One collection at a time, reviewed by a person, logged in the changelog.