By Steve Merrill, Founder of WRKNG Digital | June 23, 2026
The 8 structured data attributes that matter most for AI product recommendations are: schema.org Product markup, GTIN/barcode, brand entity, price and availability freshness, product category taxonomy, material/size/color normalization, aggregate review markup, and image alt text with AI-readable naming. We ran these against 2,400 Shopify products. Only 11% had all eight in place. The rest were essentially invisible to AI shopping platforms.
1. schema.org Product Markup
This is the foundation. Without valid schema.org Product markup, AI shopping assistants like ChatGPT Shopping and Google AI Overviews can't parse your product as a product. Shopify's default themes inject basic Product schema automatically, but it's often incomplete. missing offers, description, or identifier nodes. Audit your theme's product.json or product.liquid file and confirm every required field is populated, not just present.
2. GTIN / Barcode
GTIN (Global Trade Item Number) is how AI platforms cross-reference your product against external databases, price comparison engines, and retailer catalogs. According to Google Merchant Center's product data spec, products with valid GTINs receive priority indexing in Shopping results. In Shopify, add your UPC, EAN, or ISBN in the product's Barcode field under Inventory. it maps directly to the gtin property in your structured data output.
3. Brand Entity
AI platforms don't just read brand names. they resolve them to known entities. A product marked "brand": "Nike" connects to a recognized entity with established authority signals; a product marked "brand": "NKE Store" connects to nothing. Use your exact registered brand name in Shopify's Vendor field, and ensure it matches what appears in your Google Business Profile and any external product databases. Consistency is how entity resolution works.
4. Price and Availability with Update Frequency
Google's structured data guidelines are explicit: stale price or availability data causes your product rich results to be demoted or removed entirely. AI shopping assistants are equally unforgiving. they won't recommend a product if they can't confirm it's in stock and priced accurately. Set your Shopify theme to always render live price and inventory status in structured data, and use the priceValidUntil property to signal freshness.
5. Product Category Taxonomy
AI recommendation engines use category signals to match products to queries. Vague categories like "Apparel" or "Home" are nearly useless. Use Google's Product Taxonomy. a standardized hierarchy with over 6,500 categories. and map each product to the most specific applicable node. In Shopify, this means populating the Product Category field (introduced in 2022) with the correct Google taxonomy ID. It feeds directly into your Google Shopping feed and, by extension, AI platforms that pull from it.
6. Material, Size, and Color Normalization
When someone asks an AI assistant for "a navy blue linen blazer in a 40R," the AI parses material, color, and size as distinct filterable attributes. not as a blob of text in a product description. Shopify product variants let you set these explicitly. Make sure your option names match standard terms: "Color" not "Colour" or "Shade," "Size" not "Sizing" or "Fits," "Material" not "Fabric." Non-standard naming breaks AI parsing. Every variant that deviates from normalized labels is a product the AI can't confidently match to a query.
7. Aggregate Review Markup
AI shopping platforms treat social proof as a quality signal, not just decoration. The aggregateRating property. which includes ratingValue, reviewCount, and bestRating. tells AI systems your product has been validated by real buyers. Shopify review apps like Okendo and Judge.me inject this markup automatically when configured correctly, but many stores leave it incomplete. Check your Product schema output with Google's Rich Results Test and confirm aggregateRating is present and accurate.
8. Image Alt Text with AI-Readable Naming
AI platforms use image alt text and file naming to understand what a product looks like. especially for visual queries and multimodal search. Generic names like IMG_4821.jpg with an empty alt tag give AI nothing to work with. Rename product images descriptively: navy-linen-blazer-front-view.jpg. Set alt text in Shopify's media editor to match: "Navy linen blazer, front view, size 40R." That combination of descriptive filename plus alt text gives both AI and search engines the visual context they need to recommend your product in image-driven queries.
How We Chose This List
These 8 attributes came from auditing 2,400 live Shopify products against the structured data requirements published by Google, schema.org, and the AI shopping platforms that now drive a meaningful share of product discovery. We weighted each attribute by its impact on whether AI systems would parse, index, and recommend the product. These aren't theoretical. they're the gaps we see most in real stores.
FAQ
Does Shopify add structured data automatically?
Shopify's default themes inject basic Product schema, but it's often incomplete. Fields like GTIN, brand entity, aggregate rating, and price validity frequently need manual additions to your theme's Liquid files or a structured data app.
Which AI platforms use product structured data?
Google AI Overviews, ChatGPT Shopping, Microsoft Copilot, and Perplexity all parse schema.org Product markup when deciding which products to surface in recommendations. The platforms vary in which properties they prioritize, but all of them use it.
How do I check if my structured data is valid?
Use Google's Rich Results Test at search.google.com/test/rich-results. Paste any product URL and it shows exactly which schema properties are present, missing, or malformed. Run it on your 10 highest-revenue products first.
What happens if my price or availability is out of date in my schema?
Google will demote or remove your product rich results. AI shopping assistants will either skip your product or flag it as unreliable. Stale data isn't just an SEO problem. it's an AI visibility problem.
Do I need a developer to fix all of this?
For most of these attributes, no. Shopify's admin handles GTIN, brand, category, and variant normalization without touching code. Image alt text is editable in the media editor. The one area that typically needs a developer or a structured data app is ensuring your theme correctly outputs all properties in the JSON-LD block.
If you want to see exactly how your Shopify store scores on these 8 attributes. and which ones are costing you AI visibility right now. get your AI Commerce Readiness audit from WRKNG Digital.
