By Steve Merrill, Founder of WRKNG Digital | October 20, 2026
Why won't AI shopping assistants recommend my store?
Most of the time it's not your product. It's your feed. AI shopping assistants read structured product data, not your pretty homepage, and if that data is thin or broken you get skipped. Here are the ten fixes that get your Shopify store read, trusted, and recommended.
1. Write titles the way shoppers ask AI
Your product title is the first thing an assistant matches against a question. Lead with the brand, product type, and the one spec people search for, like "Allbirds Wool Runner Men's Sneaker Gray." According to Google Merchant Center, strong titles front-load the attributes that matter and cut the fluff.
2. Fill every required Google Merchant attribute
A feed with missing required fields gets disapproved, and a disapproved product can't be recommended by anyone. Fill id, title, description, link, image_link, price, and availability on every single item. Google's product data specification lists them all.
3. Add GTINs and brand on every product
A GTIN is the barcode number that lets an assistant match your product to a known item and trust it. Products without one often lose to a competitor the AI can actually verify. If you make your own goods, set the brand and mark the item as having no GTIN so the feed stays clean.
4. Rewrite descriptions to answer real questions
Shoppers ask AI full questions, so your description should answer them in plain sentences. State what it's made of, who it's for, and how it fits in the first two lines. Skip the slogans. The assistant is reading for facts, not vibe, and it quotes the facts back to the shopper.
5. Keep price and availability in sync
Nothing kills a recommendation faster than a price that's wrong or a product that's out of stock. Assistants drop stores that show stale data because it burns the shopper's trust. Sync price and stock at least daily, and use real-time updates during sales.
6. Use the full Google product category tree
Assign the most specific Google product category you can, not the broad parent. "Apparel > Shoes > Athletic Shoes" tells an assistant far more than "Apparel." The deeper the category, the more questions your product qualifies for. Broad buckets get you lost in a crowd of thousands.
7. Add structured attributes: color, size, material, age
When someone asks for "a waterproof jacket in large for a kid," the AI filters on structured attributes, not your description text. Fill color, size, material, gender, and age_group wherever they apply. These fields are how your product survives a narrow query.
8. Give every product a clean, high-res image
Shopping assistants show images next to the pick, and a blurry or watermarked photo gets your item passed over. Use a clear shot of the product on a plain background, at least 800 by 800 pixels. Google's image guidelines spell out what gets rejected.
9. Add Product schema on your store pages
The feed feeds shopping tools. Product schema feeds the AI crawlers that read your site directly. Mark up price, availability, and reviews using Schema.org Product so both paths tell the same story.
10. Fix the errors sitting in Merchant Center
Open the Diagnostics tab and read it. Most stores have dozens of disapproved items they never knew about, and every one is a product no assistant can recommend. Shopify's Google channel docs show how your feed connects, and Merchant Center tells you exactly what's broken.
How We Chose This List
We ranked these by what actually moves products from invisible to recommended across the live Shopify stores we run. The fixes near the top are the ones that break feeds most often and cost stores the most sales. None of this is theory. It's what we check first on every account.
FAQ
What product feed matters most for AI shopping assistants?
Your Google Merchant Center feed. ChatGPT, Gemini, and Perplexity shopping tools pull from the same structured product data Google already uses. A clean Merchant Center feed is the single biggest lever you have.
Do GTINs really affect whether AI recommends my product?
Yes. GTINs let an assistant match your exact product to a known item and trust it. Products without one often get skipped for ones the AI can verify.
How often should I update my feed?
Price and availability should update at least daily, faster if you run flash sales. Stale data is the fastest way to get dropped from a recommendation.
Does Product schema on my site matter if I already have a feed?
It does. The feed feeds shopping tools, and Product schema helps AI crawlers that read your site directly. You want both saying the same thing.
Want your feed built to get recommended instead of skipped? See how we do it at WRKNG Digital's agentic commerce page.
