By Steve Merrill, Founder of WRKNG Digital | July 28, 2026
The nine product feed fields that decide whether AI recommends your Shopify store are title, description, GTIN, brand, availability, price, product type, images, and structured attributes. Get these right and AI can read your catalog, trust it, and put your products in front of a buyer. Get them wrong and you are invisible.
I ran a $10M clothing brand for 15 years. Bad product data cost me sales I never even saw. AI does not guess. It reads the feed. Here is what to fix, field by field.
1. Product Title
The title is the single field AI weighs most, because it is the first thing that gets matched to a shopper's question. Lead with brand, then product, then the one or two attributes that matter (color, size, material). Google's own product title guidance says front-load the important words. Keyword-stuffed titles get ignored. Clear ones get picked.
2. Product Description
AI reads the description to confirm the title and answer follow-up questions like "is this waterproof" or "does it fit a toddler." Write the first sentence as a plain answer, not a marketing slogan. The Merchant Center description spec rewards accurate, detail-rich copy over hype. Say what the thing is and who it is for.
3. GTIN
The GTIN is the barcode number that ties your product to a global identity every shopping system already knows. When AI sees a valid GTIN, it can cross-reference reviews, pricing, and specs from other sources, which is trust it cannot build from your word alone. Google's GTIN documentation treats it as required for most branded goods. No GTIN, less trust.
4. Brand
Brand tells AI who makes the product, and it is how a shopper asking for a specific label ever finds you. Fill it even on your own private-label goods. A blank brand field is a dead end. When someone asks an assistant for "that brand," you want to be in the answer, not left out.
5. Availability
Availability decides whether AI recommends you right now or skips you for a competitor with stock. AI will not send a buyer to an out-of-stock page, so a stale "in stock" flag burns trust fast. Sync this in near real time. The fastest way to get dropped from an AI recommendation is showing stock you do not have.
6. Price
Price has to match what the shopper sees at checkout, or AI flags the mismatch and pulls the recommendation. Include the currency and keep sale prices current with the correct sale-price field. A price that is off by even a dollar reads as untrustworthy data. Accuracy beats being the cheapest.
7. Product Type and Google Product Category
These two fields tell AI where your product sits in the catalog, and they are how you show up for a category search instead of only an exact name. Use your own product_type for internal structure and the official Google product category taxonomy for the standardized bucket. Wrong category, wrong shelf. AI puts you next to the right competitors when the category is right.
8. Images
AI increasingly reads images to confirm a product matches the text, so a clean main image on a plain background does real work now. Blurry or watermarked shots get penalized and read as low quality. Give it a high-resolution main image and a few honest angles. The picture is data, not decoration.
9. Structured Attributes
Structured attributes are the specific fields like color, size, material, gender, and age group that let AI filter your product into the exact request a shopper makes. "Red running shoes size 10" only finds you if those attributes are filled in as data, not buried in the description. Mirror them on your live pages with Schema.org Product markup so crawlers read the same facts twice. Fill every attribute you have. Empty fields are lost sales.
How We Chose This List
These nine fields map directly to the Google Merchant Center product data specification and the Schema.org Product type, the two standards AI shopping tools read from most. I ranked them by how often a missing or wrong value knocks a product out of an AI recommendation, based on live Shopify feeds we audit at WRKNG Digital.
FAQ
What is a product feed for AI search?
A product feed is the structured data file that lists every product attribute AI shopping tools read to understand and recommend your store. It maps to the Google Merchant Center product data specification and covers title, description, GTIN, brand, price, and availability.
Do I need GTINs for AI to recommend my Shopify products?
Yes for anything with a manufacturer barcode. GTINs let AI match your product to a known global identifier, which builds trust and cross-references reviews and pricing. Skip them only for truly custom or handmade items that have no barcode.
How often should I update my Shopify product feed?
Price and availability should update in near real time, ideally through an automated sync. Stale stock or price data is one of the fastest ways to get dropped from an AI recommendation.
Does Schema.org Product markup matter if I already have a feed?
Yes. The feed powers shopping surfaces, and Schema.org Product markup on your live pages helps AI crawlers read the same data straight from the page. Keeping both in sync gives AI two consistent sources instead of one.
Your feed is the difference between AI recommending your store and never seeing it. We fix product data for AI shopping every day. See how at WRKNG Digital's agentic commerce page.
