By Steve Merrill, Founder of WRKNG Digital | June 19, 2026
The fields ChatGPT Shopping needs from your product feed are the same ones you've probably been ignoring. Title, description, GTIN, category taxonomy, attributes , most Shopify stores leave at least half of these incomplete or wrong. AI shopping assistants don't guess when data is missing. They skip your product and recommend someone else's.
Here's what complete looks like across all 10 fields.
1. Product Title
Your product title is the first signal AI uses to match your product to a shopper's query. Incomplete looks like: "Blue Shirt" or "Running Shoe v2." Complete looks like: "Men's Lightweight Trail Running Shoe , Breathable Mesh, Wide Toe Box, Sizes 8–14." Lead with the specific benefit, include key attributes, and keep it between 70 and 150 characters. Shopify's product editor makes this easy to update in bulk.
2. Product Description
AI shopping assistants pull product descriptions to answer buyer questions directly. A three-sentence blurb doesn't give them enough to work with. Write 200+ words that answer the real questions a buyer would ask , what it's made of, who it's for, what problem it solves, and what it comes with. Conversational language outperforms keyword-stuffed copy here. Think about how you'd describe the product to a friend, then write that down.
3. High-Quality Images
ChatGPT Shopping and Google AI shopping surfaces display product images in recommendations. A single dark, blurry image gets skipped. Complete means multiple angles on a clean white or neutral background, at least one lifestyle shot, and images formatted at 1:1 ratio for consistency across platforms. Google's image requirements for Merchant Center are a good benchmark even if you're not running Google Shopping ads.
4. Price and Availability
Stale pricing is one of the fastest ways to get your products downranked by AI shopping tools. If an AI assistant recommends a product at $49 and the buyer lands on a page showing $79, that's a broken experience , and AI platforms track that signal. Your feed needs real-time price and availability data, which means connecting your Shopify store directly via a live product feed rather than a static export.
5. Product Category / Google Product Taxonomy ID
AI classifies products using structured category data. If your category is blank or vague (like "Apparel"), AI can't place your product in the right recommendation context. Use the Google Product Taxonomy to find the exact ID for your product type and add it to your feed. This is one of the most overlooked fields in Shopify feeds , and one of the highest-use fixes you can make in an afternoon.
6. Brand Name
Your brand name needs to be identical everywhere: your Shopify store, your product feed, your Google Merchant Center account, and your social profiles. Inconsistent brand naming ("WRKNG Digital" vs. "Wrkng Digital" vs. "WrkngDigital") fragments your brand signal across AI knowledge graphs. Pick one version and use it without variation. It sounds obvious. Most stores still get it wrong.
7. GTIN / Barcode
The Global Trade Item Number is how AI platforms verify that your product is the same product listed elsewhere. Without a GTIN, Google Shopping won't show your product for exact-match queries, and Bing Shopping may filter it out entirely. If you manufacture your own products and don't have a GTIN, you need to apply for one through GS1. Skipping this field doesn't just hurt AI visibility , it hurts every channel you're selling on.
8. Product Attributes
Color, size, material, weight, dimensions. These aren't optional extras. AI shopping assistants use attributes to filter and match products to queries like "lightweight black backpack under 2 pounds." Incomplete looks like: color: "Black." Complete looks like: color: "Matte Black," material: "600D ripstop nylon," weight: "1.8 lbs," dimensions: "18 x 12 x 6 inches." More specific always wins. I've audited hundreds of Shopify product feeds at this point , vague attributes are the most common reason a product doesn't surface in AI recommendations.
9. Customer Reviews and Ratings
AI shopping tools weight social proof. A product with zero reviews or no structured rating data is harder for AI to recommend confidently. You need a minimum of four reviews with schema markup so AI can read them directly from your page. Shopify review apps like Judge.me and Yotpo support structured data natively , confirm your app is actually outputting the right AggregateRating schema before you assume it's working. Use Google's Rich Results Test to verify.
10. Return and Shipping Policy URL
Buyers ask AI assistants about return policies before they buy. If your return and shipping policy isn't machine-readable and linked directly from your product pages, AI can't surface it in a recommendation. Your policy page needs to exist at a stable URL, use plain language (not legal boilerplate), and be linked from every product page. This is also a trust signal , AI tools that recommend products increasingly factor in buyer experience signals like clear return terms.
FAQ
What product feed fields does ChatGPT Shopping require?
ChatGPT Shopping pulls from product feeds that include accurate titles, descriptions, prices, availability, brand names, GTINs, and product images. There's no published official spec, but the fields that drive recommendation quality align closely with Google Merchant Center's required and recommended attributes. Start there.
Does Shopify automatically generate a product feed for AI shopping?
Shopify generates a basic product feed, but "basic" is the problem. Default Shopify feeds often have missing GTINs, vague categories, short descriptions, and incomplete attributes. You need to actively complete those fields , the feed itself is just the delivery mechanism.
How often should I update my product feed?
For price and availability, real-time or at minimum daily. For content fields like descriptions and attributes, update whenever you make product changes. A stale feed with wrong prices is worse than no feed at all , AI platforms use feed accuracy as a quality signal.
Do product reviews actually affect AI shopping visibility?
Yes. AI shopping assistants factor in structured review data when ranking products for recommendations. A product with 50 reviews and a 4.7-star rating will outrank an identical product with no review schema, even if the content fields are the same. Schema markup is what makes your reviews machine-readable.
What's the fastest way to fix my Shopify product feed for AI visibility?
Start with GTINs and Google Product Taxonomy IDs , those are the two fields most commonly missing and highest impact. Then audit your product titles and descriptions for specificity. Run your feed through Google Merchant Center's diagnostics to surface errors. Fix the errors. That alone will move the needle.
Get Your Store AI-Ready
Most Shopify stores are invisible to AI shopping assistants right now. Not because the products are bad , because the data is incomplete. Fixing your product feed is the single highest-use thing you can do to show up in ChatGPT Shopping, Perplexity, and Google AI Overviews before your competitors do.
See how WRKNG Digital audits and improves Shopify stores for AI shopping visibility →
