By Steve Merrill, Founder of WRKNG Digital | September 11, 2026
Most Shopify stores have a product feed that technically works and practically hides them. It passes Google's checks. It just doesn't say anything an AI assistant can use.
I ran an ecommerce brand to $10M a year. Our feed was clean by every old standard, and completely silent to the new one. That gap is where most stores are sitting right now.
Here's how to fix the feed so AI search can actually read it.
Why does AI search ignore a feed that Google accepts?
Google Shopping accepts a feed if the required fields are filled. AI search reads a feed to answer a question. Those are not the same bar.
A feed can pass every rule in the Google Merchant product data specification and still have a title like "Blue Shirt - SKU 4471" and a blank material field. Accepted, and useless to a model deciding what to recommend.
Fix the title first
Your product title is the single most-read field. Write it the way a buyer describes the thing out loud, not the way your PIM names a variant.
"Merino Wool Base Layer, Men's, 200gsm" beats "Base Layer 4471 Blue M." The first is an answer. The second is a warehouse label. An assistant quotes the first and skips the second.
Fill the fields you left blank
Material, color, size, age group, gender, pattern. Most feeds leave half of these empty because Google didn't force them. AI search reads every one to match a query.
When someone asks for "a wool sweater for a cold office," the model filters on material and use. If your material field is blank, you're not in the set. Fill them, then confirm the values are real against the Google Merchant Center feed rules.
Make the description say something checkable
Feed descriptions are often the marketing blurb copied over. Rewrite them to carry facts: what it is, what it's for, who it fits. One checkable claim beats three adjectives.
A model can verify "machine washable, fits waist 30 to 40." It can't verify "effortless everyday style." Give it the first kind and it has a reason to name you.
Line the feed up with the page and the schema
Your feed, your product page, and your Product schema should all state the same facts. When they agree, a model trusts the claim. When they disagree, it treats the product as unreliable and moves on.
Run the page through the Schema.org validator and check the values match the feed. This is the read we run in a WRKNG audit: where your feed, your pages, and your data stop telling the same story.
Further reading
Frequently Asked Questions
Why does AI search ignore my Shopify product feed?
Because a feed that passes Google Shopping checks is not the same as a feed AI can read. Google accepts filled required fields; AI search reads titles, materials, and descriptions to match a question. Warehouse-style titles and blank attribute fields leave you out of the answer set.
What feed fields matter most for AI search?
Title first, then the attribute fields most stores leave blank: material, color, size, age group, gender, and pattern. AI assistants filter on these to match specific queries, so an empty field usually means you're excluded from that result.
Should my feed match my product page?
Yes. Your feed, your product page copy, and your Product schema should state the same facts. When they agree, a model trusts the claim. When they conflict, it treats the product as unreliable and recommends a store whose data is consistent.
How do I write a product title AI will read?
Write it the way a buyer describes the item out loud, with the key specs included. 'Merino Wool Base Layer, Men's, 200gsm' reads as an answer; 'Base Layer 4471 Blue M' reads as a warehouse label and gets skipped.
Want to know if AI assistants can actually find and recommend your store? Get a free AI-visibility read on your Shopify store at WRKNG Digital. We show you exactly what ChatGPT, Perplexity, and Google AI see when they look at your products.
