By Steve Merrill, Founder of WRKNG Digital — October 19, 2026
AI assistants recommend the product they understand best, not the one with the best marketing. Nine fields in your Shopify product feed decide whether ChatGPT, Gemini, or Perplexity picks you or your competitor, and most stores fill in three of them.
I've run this check in more than 40 Shopify audits. The pattern is the same every time. Great products. Empty feed fields. The AI has nothing concrete to grab, so it recommends the brand that filled theirs out.
Here's what I mean. An assistant answering "best waterproof hiking boots under $150" isn't reading your homepage copy. It's reading structured data: title, price, availability, GTIN, reviews. Fill those and you're in the running. Leave them blank and you're invisible.
How We Chose This List
I ranked these nine fields by how heavily AI product recommendation actually weighs them. The weighting comes from three things: what Google's Merchant Center requires, what Schema.org Product markup supports, and what the ChatGPT shopping feed asks merchants to submit. Fields that appear in all three, and that an AI needs to answer a shopper's question, sit at the top. Nice-to-have fields sit lower.
1. Product Title
The title is the first thing an AI reads, and it decides whether your product even matches the question. A good title names the product, the brand, and the one attribute shoppers search on. "Columbia Newton Ridge Waterproof Hiking Boot, Men's, Brown" beats "Newton Ridge Boot."
The common mistake: stuffing the title with marketing words like "best-selling" or "must-have." The AI strips those and ends up with less to match. Write the title the way a shopper describes the product to a friend.
2. GTIN and Unique Identifiers
A GTIN is the barcode number that identifies your product everywhere on earth. When an AI sees a valid GTIN, it can confirm the product is real and pull in reviews, prices, and specs from every other source selling it. Google requires GTINs for most branded products for exactly this reason.
Most Shopify stores leave this field blank. That's a mistake. Without a GTIN, the AI treats your listing as an island and can't verify a thing. Add the barcode to every branded product variant.
3. Product Type and Google Category
This field tells the AI what shelf your product belongs on. There are two parts: your own product_type and the standardized Google product category. The standardized one matters most because it maps your item to the same bucket every other retailer uses.
Stores skip the Google category or pick one that's too broad, like "Apparel" for a specific running sock. Too broad means you compete against everything. Pick the deepest, most specific category that fits.
4. Price and Availability
Price and stock status are the two fields AI checks before it recommends anything. An assistant won't suggest a $200 item for a "under $150" question, and it won't recommend something out of stock. Both fields have to be accurate and current.
The mistake I see: stale availability. The feed says "in stock" while the product page says sold out. The AI catches that mismatch and drops you. Keep availability synced in real time, and include sale prices in the sale_price field, not jammed into the title.
5. High-Resolution Images
AI shopping results are visual now, and the image is often the first thing a shopper sees next to your name. A clean, high-resolution main image on a white background wins the click. Blurry or cropped photos get skipped even when the product is a great match.
Shopify stores often upload one lifestyle shot and stop. Add a clear primary image plus additional angles through the additional_image_link field. Google's image rules are a good baseline for what AI expects.
6. Description and Structured Attributes
The description is where the AI learns the details that answer specific questions: material, size, fit, compatibility, waterproof rating. The more real attributes you spell out, the more questions your product can win. Vague descriptions win nothing.
Don't write a paragraph of brand story here. The AI wants facts. List the attributes plainly, and fill the dedicated fields Shopify offers, like color, size, material, and age_group, so the data is machine-readable, not buried in prose.
7. Brand
The brand field tells the AI who makes the product, and it connects your listing to everything known about that maker. Shoppers ask for brands by name constantly, and a missing brand field means you never show up for those questions.
This one's simple and still gets skipped. Fill the brand field on every product, even your own private-label items. Your store name is a brand too, and the AI needs it spelled out to recommend you by name.
8. Reviews and Ratings
Review data is how AI decides between two products that otherwise match. A product showing a 4.6 average across 400 reviews gets picked over an identical one with no rating attached. The assistant reads aggregate rating and review count straight from your structured data.
Most Shopify stores collect reviews but never expose them as review structured data. The stars show on the page but not in the feed. Connect your review app's output to Product schema so the AI can actually read it.
9. Shipping and Returns
Shipping cost, delivery speed, and return policy are becoming deciding factors as AI assistants move from recommending products to helping people buy them. A shopper asking "which store ships fastest" needs these fields populated to get your store in the answer.
This is the most-ignored field on the list. Add shipping details and a clear return window to your feed and your Product schema. As agentic checkout grows, the stores with this data filled in will be the ones AI can actually transact with.
FAQ
Which product feed field matters most for AI recommendations?
The product title and a valid GTIN. The title tells the AI what the product is. The GTIN lets it confirm the product is real and match it against reviews and price data from everywhere else it's sold.
Do I need a GTIN for every product to get recommended by AI?
For branded, manufactured products, yes. GTINs let AI assistants cross-check your listing against the same product everywhere else. Custom or handmade items without a GTIN should fill the brand and MPN fields instead.
Does Shopify send these fields to AI assistants automatically?
Shopify exposes Product schema and a Google-format feed, but it won't fill every field for you. GTIN, product category, and detailed attributes are often blank unless you add them to each product by hand.
How do reviews affect whether AI recommends my store?
AI assistants pull aggregate ratings and review counts when they compare options. A product with a visible 4.6 rating across 400 reviews gets recommended over an identical product with no review data attached.
Want to know which of these nine fields your store is missing right now? That's the first thing we check. See how we get Shopify stores recommended by AI at WRKNG Digital's agentic commerce page.
