Theme development, Module 5, lesson 4, 7 min

Make your products readable to AI agents

From

Priya Nair

Founder & CEO

Thursday, 8:40am

AI agents are starting to shop our catalog

Buyers are shopping with AI agents now, and Shopify's new Catalog API lets those agents query our products directly.

I want Northwind to show up when an agent searches "fruity light-roast Ethiopian beans." How should we set up our product data?

Priya

How this works

The Catalog API lets AI agents query product data directly. It reads structured data: the standard product fields plus your metafields. It does not scrape your theme's HTML.

So "will an agent represent my product correctly" is the data-modeling question from earlier in this module, asked by a new reader. Real metafields (roast level, origin, tasting notes) are what an agent can match on.

Why this matters in production

Shopify's Spring '26 Edition shipped a Catalog API that turns the product catalog into queryable infrastructure for AI agents, with image and multi-modal search. Shopify reports agent searches on the structured catalog convert at roughly twice the rate of those built on scraped data. Whether your products show up correctly in an agent's answer now depends on how their data is modeled.

Report a bug or suggestion

Architectural decision

Pick how to make Northwind's products readable to AI shopping agents.

The scenario

Priya wants Northwind's coffees to surface accurately when an AI agent searches "fruity light-roast Ethiopian beans under $25."

Right now the roast level, origin, and tasting notes live only in the product description and the theme's PDP markup.