By Steve Merrill, Founder of WRKNG Digital | September 27, 2026
The feed fields that decide AI recommendations are GTIN, brand, price, availability, and specific attributes like material and condition. ChatGPT and Perplexity match these against a shopper's request. Miss one and you drop out of the answer. Here are the nine that matter most.
1. GTIN
The global trade item number is how an assistant confirms your product is the real thing. A missing GTIN reads as an unverified listing, and unverified products get skipped. This is the single most common gap we find in Shopify feeds.
2. Brand
Assistants use brand to match "best [brand] running shoes" style requests and to establish trust. A blank brand field makes your product ambiguous. Fill it on every SKU, even for house brands.
3. Price and currency
ChatGPT and Perplexity won't recommend a product they can't price. Include the exact price and currency, and keep it synced. Stale or missing price data reads as unreliable.
4. Availability
An assistant won't send a shopper to something out of stock. Accurate, real-time availability keeps you eligible. Google's Merchant Center spec treats this as required, and AI assistants lean on it the same way.
5. Product title
The title is prime matching text. Lead with the specific product, not clever branding. "Merino wool base layer, men's, long sleeve" beats a cute name every time for AI matching.
6. Product type and category
Category fields tell the assistant where your product fits. Correct product type and Google product category help you appear for the right requests and avoid mismatched ones.
7. Material and specific attributes
Attributes like material, color, size, and compatibility are what assistants match against detailed queries. "Waterproof, leather, size 10" only works if those attributes exist in the feed. This is where metafields earn their keep.
8. Image link
A valid, high-quality image link matters because assistants increasingly show product images in shopping answers. A broken or missing image can suppress your listing in visual results.
9. Condition
New, used, or refurbished. Condition is required in most feed specs and helps assistants filter correctly. Leaving it blank creates ambiguity that can cost you the recommendation.
How We Chose This List
These nine come straight from the feed audits we run on live Shopify stores. They're the fields most often missing and most directly tied to whether an assistant can confirm and recommend a product.
FAQ
Q: Which feed field is most important for AI recommendations?
GTIN. It's the primary way an assistant verifies your product is real, and it's the most common missing field we see.
Q: Are Shopify feed fields the same as Google Shopping fields?
Largely yes. The same structured fields feed Google Shopping and AI assistants, so cleaning your feed helps both channels at once.
Q: How do I add attributes like material to my Shopify feed?
Use product options and metafields to store specific attributes, then map them into your feed. That gives assistants precise details to match.
Q: Will fixing these fields help with Perplexity and ChatGPT equally?
Yes. Both read similar structured product data, so complete, accurate fields improve visibility across assistants.
Want a field-by-field read on your Shopify feed? Get an audit at WRKNG Digital's agentic commerce readiness page.
