Which structured data fields make AI recommend your products?
AI recommends products with complete, consistent Product schema. Ten fields do most of the work. Get them right and you're eligible to be named. Leave them blank and you're invisible.
We've run thousands of products through audits. The pattern is always the same. Stores fill in name and price, then leave everything else empty. AI reads the gaps as risk. Here are the ten fields that matter.
1. name
The exact product name AI will speak back to a shopper. Keep it clean and specific. No keyword stuffing, no marketing fluff crammed into the title.
2. description
A real description with materials, use case, and fit. This is what AI quotes. Vague copy gets skipped for a competitor who wrote like a human.
3. brand
Your brand, stated consistently everywhere. AI uses this to connect your product to your business as one entity. Inconsistency here quietly erases you.
4. gtin / mpn
Global identifiers let AI match your product across the web and trust it's real. Missing GTINs are one of the most common gaps we find, and one of the most damaging.
5. offers price and priceCurrency
Price and currency, structured, not just printed on the page. AI needs the machine-readable value to compare and recommend. See the schema.org Offer spec.
6. availability
In stock or out, kept accurate in real time. Stale availability is a trust killer. If the data lies once, AI stops quoting you.
7. aggregateRating
A summary rating from real reviews. This is social proof AI can read. A product with ratings beats an identical product with none, every time.
8. review
Individual review markup gives AI quotable proof from actual buyers. It's the difference between "this exists" and "people like this."
9. image
A valid, high-quality image URL in the schema. AI-driven shopping surfaces increasingly show product images. A broken or missing image URL drops you from visual results.
10. itemCondition and shippingDetails
Condition and shipping data round out the record. As agentic commerce grows, AI needs to know it can actually complete the purchase. Google's docs now weight shipping and returns data more heavily.
Where to start
Here's the bottom line: completeness beats cleverness. A boring, fully populated record wins over a beautiful page with half the fields blank.
Audit every field across your catalog. The gaps are where your competitors are beating you in AI answers right now.
Frequently asked questions
What data does AI use?
Product schema: name, brand, GTIN, price, availability, and ratings.
Which field matters most?
None alone. AI trusts complete, consistent records over any single hero field.
Does Shopify add this automatically?
Partially. Coverage varies by theme and app, and most stores have real gaps.
Get a WRKNG Digital structured data audit.
By Steve Merrill, WRKNG Digital. Published August 22, 2026.
