Shopify product data that AI needs to find your product

One of the key takeaways from Shopify's AI discovery session at DotDev was that the product spec now answers the query. Get the category wrong or skip it and that product becomes invisible to the exact query that would have converted.

Same shift in different platforms

Amazon's Alexa, Home Depot's assistant, and Sephora's chat interface are all nudging shoppers to ask questions rather than type keywords.

None of that works without structured data underneath to match the question against. Shopify pointed to this as evidence the shift is happening industry-wide.

I broke down the 7 checks from that DotDev session here

From keyword to use case

Earlier - in a usual shopping journey - on a storefront, a shopper used to type "shorts" into a search bar and browse until something looked right.

That's not how it works anymore.

Now when someone asks ChatGPT or Gemini for shorts, they're starting from a reason, not a word - swim shorts, running shorts, something for a hot commute.

The query already carries the use case.

The product either answers it or it doesn't show up. Whether that listing surfaces for a swim query or a running query has nothing to do with the title or the photos.

It comes from category metafields. Those only exist because of a fixed mechanism underneath - assign the right category to a product. And Shopify opens the exact attribute set that matters for it: material, closure, intended use, fit, for a pair of shorts. Skip the category or get it wrong, and none of those fields exist to fill in.

What missing metafields cost you

Two of the nine category fields are filled in: color and target gender. The other seven, including fabric, fit, and care instructions, are empty.

For a query, none of the fields that would answer it exist yet. However well the product fits, there's nothing there for that query to match against.

Brands sit exactly like this by default. Filling these in across hundreds or thousands of SKUs is the part almost nobody gets to - the biggest visibility gap.

Same product, run through the Atomz AI. Fabric, care instructions, and clothing features are filled in. AI reads every product it touches instead of guessing.

The same fields power your AI discovery

This isn't limited to the agent shopping or external AI channels. Your on-site search and assistant read from the same enriched catalog on your Shopify admin.

It's the same mechanism that Ulta, Amazon, Walmart are all building.

Check out what AI shopping assistants run on →

Pull up your best-selling products in Shopify admin and check whether the category-specific fields are populated, beyond the title and description. Most aren't.

Run that check for your whole catalog →

If you'd rather have it filled in automatically, Atomz AI does this best.

- Ankit