By Steve Merrill, Founder of WRKNG Digital | September 22, 2026
AI shopping assistants read your schema before your copy. Eight fields do most of the deciding. Fill them and you're a candidate. Leave them blank and you're a guess.
Here's the list, in the order that matters.
1. Price
The single most-checked field. It belongs in your Offer markup as a number with a currency. If it loads by script after the page, a machine may read it as blank. Put it in the raw HTML.
2. Availability
In stock or out. An assistant won't recommend a product it can't confirm you can sell. Missing availability is the most common reason ChatGPT hedges with "check the site."
3. GTIN or MPN
The unique product identifier. It's how an assistant matches your item to the wider web and its reviews. Google's data spec treats it as strongly recommended, and AI shopping uses the same signal.
4. Brand
Ties the product to a maker the assistant may already trust. A missing brand field turns a known product into an anonymous one, and anonymous products lose to named ones.
5. Product name and description
The Product name and description give the assistant the words to match a query. Write real specs here, not marketing fluff. The machine quotes facts, not adjectives.
6. Image
A valid, direct image URL in the schema. Assistants increasingly show product images in results, and a broken or missing image URL can drop you from the visual answer entirely.
7. Review and rating markup
Review and aggregate rating data give the assistant social proof it can cite. Stores with structured reviews get recommended more often because the assistant has something concrete to say about them.
8. Shipping and return details
Shipping cost and return terms as structured data. Agentic checkout weighs these directly. A store that hides shipping until checkout looks like risk to a machine comparing options.
How We Chose This List
These eight map to the fields AI shopping tools and the underlying product data specs actually read, drawn from Google's Merchant spec and how ChatGPT and Perplexity describe their shopping results. We ordered them by how often a missing value causes an assistant to skip a product.
Further reading
- Schema.org Offer reference
- Schema.org Product reference
- Google Merchant Center product data spec
- Google product structured data guidelines
- Schema.org Review reference
Frequently Asked Questions
What schema fields matter most for AI shopping?
Price and availability lead, followed by GTIN or MPN, brand, product name and description, image, review markup, and shipping and return details. AI assistants check these before recommending a product, and a missing value in any of the top few is enough to make an assistant skip your item.
Do I need GTIN on every Shopify product?
For any product that has one, yes. The GTIN lets an assistant match your item to the wider web, its reviews, and price comparisons. Custom or handmade products without a GTIN can use MPN and brand instead, but standard retail products should always carry the GTIN.
Does schema replace a product feed?
No, they work together. Schema lives on your live product page and a feed is sent to platforms like Google Merchant Center. AI assistants pull from both, so keep them consistent. Conflicting price or availability between the two makes an assistant trust your store less.
Want to know if AI assistants can actually find and recommend your store? Get a free AI-visibility read on your Shopify store at WRKNG Digital. We show you exactly what ChatGPT, Perplexity, and Google AI see when they look at your products.
