Google Shopping feed requirements: every field that matters in 2026

Google Shopping requires seven attributes for every product: id, title, description, link, image link, availability and price. Brand, GTIN or MPN, condition and, for apparel, color, size, gender and age group become required in many cases. Meeting the requirements gets a product approved. The optional fields decide which searches it can match.

One sporting goods retailer's product descriptions looked fine. Every fact was correct, but every fact showed up twice: once in the copy, and again underneath as a list of colon-separated specs. Nobody did this on purpose. The feed build was merging two fields, the product description and the product highlights, into one, so Google got one long block of text with no structure. When Lily Max audited 20 of those products, moving each fact back into its own field took 344 attribute changes, including 226 entries in the product detail field.

That feed met every Google Shopping requirement, but most of what a shopper might ask about was stuck inside one paragraph. So this guide covers two things: the fields Google requires, and the fields that decide whether an approved product gets found. Every requirement below comes from Google's product data specification.

Which fields does Google Shopping require for every product?

Seven. Google says that if a required attribute is missing, the product “won't be able to serve in ads and free listings.”

The seven attributes Google requires for every product, with Google's limits and the gap we see most often.

When do brand, GTIN and MPN become required?

It depends on what you sell. Here's what Google's specification says:

Conditional requirements in Google's product data specification.

What extra fields does apparel need?

Four, in the six markets Google lists: Brazil, France, Germany, Japan, the UK and the US.

Apparel requirements in Google's product data specification.

Whether these apply depends on the product’s category, which Google assigns automatically (you can override it). If Google files a shirt outside apparel, it can miss requirements it should meet. Our guide to product taxonomy covers how categorization decides what a product is eligible for.

What are Google's image requirements?

Images need to be crawlable, accurate and free of promotional overlays. Google accepts JPEG, WebP, PNG, non-animated GIF, BMP and TIFF. Don't upscale images, and don't submit thumbnails, placeholders, generic images, promotional text, watermarks or borders. You can add up to 10 more images showing the product in use.

Two changes to plan for now:

• A higher minimum size. Google's specification announces a new minimum of 500 x 500 pixels for all product images, with enforcement beginning 31 January 2027. If your catalog has older, smaller images, you have until then to replace them.
• AI-generated images keep their labels. Images made with generative AI must keep their embedded metadata, such as the IPTC DigitalSourceType tag.

Which optional fields decide what a product matches?

The ones a shopper's question can be checked against. Google's specification now has several fields made for this:

Optional Google Shopping attributes that carry product facts, with Google's limits.

The sporting goods retailer from the start of this post already had every fact a shopper would need. It just didn't have each one in its own field. Specs belong in product highlights and product details, and Google's rules for both say not to repeat what's already in the description. So if your description repeats its specs underneath, or your feed merges two fields into one, you're working against how Google set these fields up.

For which attributes matter most by category, see product attributes. For the fields Google added for AI shopping surfaces, see Merchant Center conversational attributes.

Does an AI-written title need a different field?

Yes. Google's specification has separate attributes for text created with generative AI: structured_title and structured_description. Each takes the content (up to 150 characters for a title, 5,000 for a description) and an optional digital_source_type that marks it as algorithmically generated. Anything written without generative AI goes in the regular title and description. If you're using AI to write product copy across your catalog, check which field it's going into.

A Google Shopping feed requirements checklist

1. 1.Confirm all seven required attributes are present for every product, and that IDs stay stable between builds.
2. 2.Check brand, GTIN, MPN and identifier_exists against what each product actually has.
3. 3.For apparel in the six listed markets, fill color, size, gender, age group and item group ID.
4. 4.Compare feed price and availability against the landing page during your next sale.
5. 5.List any images under 500 x 500 pixels and replace them before 31 January 2027.
6. 6.Split specs out of descriptions into product_highlight and product_detail, and say each fact once.
7. 7.Add question_and_answer pairs for the questions your customer service team hears most.
8. 8.Write product_type paths that carry the detail Google's taxonomy lacks, with the first value set for bidding.
9. 9.Route AI-generated titles and descriptions to structured_title and structured_description.
10. 10.Re-read Google's specification each quarter. Requirements change, and the image rule above shows how much notice Google gives.

What we don't know yet

Meeting every requirement gets your product approved. It doesn't guarantee impressions, clicks or sales, and Google doesn't say how much each optional field counts in matching. The retailer at the top of this page had a compliant feed before any change was made. The only way to know if a restructured record earns more is a controlled test, which is how a feed change should be judged before it is rolled out.

What this means for your team

Most teams stop working on Merchant Center once a product is no longer disapproved. But Google's requirements are the minimum. A feed can pass all of them and still keep its best facts in one long description. The fields to work on next are the ones no error report will ever flag.

Lily Max goes through your Merchant Center feed field by field, puts each fact where Google reads it, and tests the changes against a control. Book a demo to see it in your catalog.