By Steve Merrill, Founder of WRKNG Digital | October 25, 2026
The product feed fields that decide whether AI recommends your Shopify products are title, description, GTIN/MPN, brand, product category, price, availability, image, and attributes like color and size. AI shopping assistants read these fields to match products to a shopper's request, and a thin or messy feed gets skipped. Here is what each field does and why it matters.
I ran a $10M ecommerce brand for 15 years. The feed was always the boring part. It is not boring anymore. It is the thing that gets you picked.
1. Product Title
The title is the first thing an AI assistant reads to decide what your product even is. It should lead with the brand and the exact product type, then the key attribute a shopper would type. A title like "Allbirds Wool Runner Sneakers, Men's, Gray" gives the model more to match against than "Comfy Shoes." Google Merchant Center recommends front-loading the most important terms.
2. Description
The description is where AI pulls the details the title cannot hold. Material, fit, use case, and what makes it different. Write it for a human asking a specific question, because that is what the assistant is doing on their behalf. Keyword stuffing hurts you here. Clear sentences win.
3. GTIN / MPN
The GTIN (barcode) and MPN (manufacturer part number) are how AI knows your product is the same real-world item sold elsewhere. These unique IDs let the model compare your price and reviews against every other listing of that exact product. Missing them makes you look like an unknown item, and unknown items get recommended less. Google's GTIN guidance covers which products need them.
4. Brand
The brand field tells AI who makes the product, which matters because so many shoppers ask for a brand by name. If someone asks an assistant for "Nike running shorts," a blank brand field drops you out of that answer entirely. Fill it on every product, even your own private label.
5. Google Product Category and Product Type
Category fields put your product in the right aisle so AI recommends it for the right request. Google Product Category uses a fixed taxonomy, and your own product type field adds a more specific path. Get the category wrong and your running shoe shows up for someone shopping for dress shoes. Nobody buys that.
6. Price and Sale Price
Price is a hard filter for most AI shopping queries, because people shop with a budget in mind. The feed price has to match your live Shopify price exactly, or the assistant flags a mismatch and trust drops. Keep the sale price field accurate too, since "on sale" is a common thing shoppers ask for.
7. Availability
Availability decides whether AI even bothers recommending you, because no assistant wants to send a shopper to an out-of-stock page. This field has to update in near real time as inventory moves. An agent that is about to place an order needs to know the item is actually there. Stale availability is a fast way to get dropped.
8. Image Link
The image is how both the shopper and the AI confirm the product is what they think it is. Use a high-resolution shot on a clean background, no watermarks, no promo text baked in. Shopify's product media guide covers the specs. A broken or low-quality image link can disqualify the whole listing.
9. Product Attributes (color, size, material, age group, gender)
Attributes are the fields that match a product to a specific request, like "blue cotton shirt, men's, large." The more of these you fill accurately, the more precise questions your product can answer. OpenAI's work on buying inside ChatGPT shows assistants lean on structured attributes to narrow choices for a shopper. Empty attribute fields mean you only show up for vague searches.
How We Chose This List
We picked these nine based on which fields AI shopping assistants actually read when matching products to a request, drawing on Schema.org Product properties and Google Merchant Center's required and recommended specs. These are the fields that decide inclusion, not nice-to-haves.
FAQ
Q: What is the most important product feed field for AI recommendations?
The title and GTIN matter most. The title tells AI what your product is, and the GTIN proves it is a real, comparable item sold across the market.
Q: Do I need a GTIN if I make my own products?
For private-label products you manufacture yourself, a GTIN is often not required, but you should still fill the brand and MPN fields. That gives AI enough to identify the item.
Q: How often should availability and price update in my feed?
As close to real time as you can manage. AI assistants penalize feeds where price or stock does not match your live Shopify store, because a mismatch breaks shopper trust.
Q: Does Shopify generate a feed AI can read automatically?
Shopify can produce a product feed, but the default fields are often thin. You usually need to map and enrich the feed so every field above is filled and accurate.
Get Your Feed Ready for AI Shoppers
If AI assistants are picking products for your buyers, a clean feed is the difference between getting recommended and getting skipped. We get Shopify stores ready for this. See how at WRKNG Digital's agentic commerce page.
