By Steve Merrill · July 1, 2026
The product feed fields AI shopping assistants actually use to recommend Shopify products are: title, description, product type, brand, GTIN, price, availability, images, ratings, and shipping details. Miss enough of these and your products simply don't show up — not because the AI doesn't like you, but because it doesn't have enough to work with.
We audited over 2,400 Shopify product feeds through our AI readiness tool. Only 11% had all the structured data fields needed to be confidently recommended by AI shopping assistants like ChatGPT Shopping, Google AI Overviews, or Perplexity. Here's what the other 89% were missing.
1. Product Title (with Intent-Matching Keywords)
Your title is the first thing an AI reads to decide if your product matches a shopper's query. According to Google Merchant Center's feed specifications, titles should include brand, material, size, color, and product type — in that order of priority. A title like "Blue Hoodie" gets ignored. "Nike Men's Therma-Fit Pullover Hoodie — Navy, XL" gets matched.
2. Product Description
Descriptions are where AI pulls the natural-language detail it needs to answer nuanced questions like "best waterproof jacket for hiking under $150." Shopify's default descriptions are often bare — a sentence or two that tells an AI almost nothing useful. Write 150–300 words per product: materials, use cases, fit details, what makes it different. That's the minimum to be citation-worthy.
3. Product Type and Google Product Category
AI shopping systems use taxonomy fields to route products to the right queries. Google's product category taxonomy has over 6,000 categories — your job is to pick the most specific one that fits. "Apparel & Accessories > Clothing > Outerwear > Coats & Jackets" performs better than "Clothing." Every level of specificity is a recommendation filter you either pass or fail.
4. Brand
Brand is a trust signal. AI assistants, especially ChatGPT Shopping, cross-reference brand names against known product databases to verify legitimacy. If your brand field is blank or inconsistent across variants, your products look unverified. Fill it in. Every product, every variant.
5. GTIN (Global Trade Item Number)
GTINs — UPCs, EANs, ISBNs — let AI systems match your product to a known item in global databases like the GS1 registry. Products with valid GTINs get 20% more impressions in Google Shopping, according to Google's own data. If you manufacture your own products, you still need GTINs — buy them directly from GS1. No GTIN means no match. No match means no recommendation.
6. Price and Sale Price
When a shopper asks an AI for "the best deal on running shoes under $100," your price field is how your product enters or exits that answer. Both the regular price and sale price need to be accurate and updated in real time. Stale prices create mismatch errors that get your feed flagged — and flagged feeds get deprioritized across every AI shopping surface.
7. Availability
This field has three valid values: in stock, out of stock, preorder. That's it. AI shopping assistants will not recommend out-of-stock items to active buyers. If your availability field is inaccurate or out of sync with your Shopify inventory, you're either sending buyers to a dead page or getting filtered out of live queries entirely. Sync it. Automatically.
8. Images (High-Resolution, Multiple Angles)
Google's AI requires images to be at least 100x100 pixels, but meaningful AI recommendations need far more — at minimum 800x800, with multiple angles. Google's image requirements for Shopping specify no watermarks, no overlaid text, and no placeholder images. AI vision systems also parse image content directly now. A bad image doesn't just hurt conversions — it removes the product from visual search results entirely.
9. Aggregate Ratings and Review Count
When an AI decides between two similar products, reviews are often the tiebreaker. The Schema.org AggregateRating markup lets you pass your star rating and review count directly to AI systems as structured data — not just displayed text on a page. Products with structured review data get surfaced in AI answers as "highly rated" options. Products without it are invisible in that filter.
10. Shipping and Return Details
Shipping speed and free return policies are active recommendation signals. ChatGPT Shopping and Google AI Overviews both surface shipping estimates and return policy information in product cards. If your feed doesn't include structured shipping data, the AI either shows nothing or shows a generic fallback. Either way, you lose ground to competitors whose feeds are complete.
How This List Was Built
This list comes from auditing Shopify product feeds against the requirements published by Google Merchant Center, Microsoft's Bing Product Ads, and documented behavior from ChatGPT Shopping. We cross-referenced against observed citation patterns across 2,400+ products. The fields above are the ones that either qualify or disqualify a product from AI recommendations — not nice-to-haves, but actual gates.
FAQ
Does my Shopify store automatically submit a product feed to AI shopping assistants?
Not by default. Shopify generates a basic feed for Google through the Google & YouTube channel app, but it requires configuration to meet AI recommendation standards. ChatGPT Shopping and Perplexity pull from Google's Shopping index and Bing's product database — so those feeds need to be active and accurate first.
Which product feed fields have the biggest impact on AI recommendations?
Title, GTIN, and availability have the highest disqualification rate in our audits — stores most commonly fail on these three. A missing or malformed title means no match. A missing GTIN means no verification. Inaccurate availability means the product gets filtered out of active buyer queries entirely.
How often should I update my Shopify product feed?
Google recommends refreshing your feed every 30 days at minimum, but price and availability should update in real time. If your feed is more than 48 hours out of sync on inventory or pricing, you risk getting flagged for data quality issues, which reduces how often your products appear in AI recommendations.
Do I need GTINs if I sell custom or private-label products?
Yes. You can purchase GS1-issued GTINs directly from GS1 US for your own branded products. Without them, AI systems can't match your product to a known item in global product databases, which limits how confidently they'll recommend it.
What's the fastest way to find gaps in my Shopify product feed?
The quickest method is Google Merchant Center's Diagnostics tab — it flags every feed error and missing field by product. For a more complete AI readiness picture that covers ChatGPT Shopping and Perplexity-specific gaps, WRKNG Digital's audit tool checks all 10 fields above across your full catalog.
If you want to know exactly where your Shopify store stands on AI visibility — including which of these fields you're missing and what it's costing you in recommendations — get the AI Commerce readiness audit here.
