FoundRateThe Shopify AI Visibility Checklist: 12 Fixes That Decide If ChatGPT Can Recommend You
When we scanned 82 Shopify stores, 87% were never recommended by ChatGPT for their own category's buying questions. The difference between found and invisible almost always came down to the items below. Work through them in order; most stores can clear the whole list in an afternoon, and none of it needs a developer.
1. Every product has a real description (200+ characters of substance)
State what it is, what it's made of, its size or dimensions, and who it's for, in plain sentences. In our scan data, 19% of products had descriptions too thin for an AI to work with. "You'll love this!" is not information; "12cm stoneware pour-over dripper, fits #2 filters, dishwasher safe" is.
2. Product type is set, and specific
24% of products we scanned had no product type at all. Most AI buying questions are category questions ("best ceramic dripper", "top rated beard oil"), and the product type is the strongest category signal in your catalog. Avoid vague types like "Misc", "Accessories" or "New"; use "Coffee Dripper", "Beard Oil".
3. A real price on every variant
Budget questions ("under $50") match on price. Zero-dollar placeholders and broken variant prices silently exclude you from every one of them.
4. Three or more images per product
38% of products we checked had fewer than three. Sparse imagery reads as low-trust to both shoppers and the systems ranking products for them.
5. Vendor / brand field is filled in
The AI attributes products to brands. If your own products carry no brand, there is nothing to recommend by name.
6. In stock, or hidden
AI assistants avoid recommending dead ends. Long-term out-of-stock products drag the catalog; restock them or unpublish them.
7. Product JSON-LD is intact
Structured data is how a machine verifies your product's name, price, availability and brand without guessing. Most Shopify themes emit it; some apps break it. Check yours in two minutes with this guide.
8. AI crawlers are not blocked
Check your robots.txt for rules blocking GPTBot, OAI-SearchBot, ClaudeBot or PerplexityBot. Some "SEO" and firewall apps block them by default, which is choosing invisibility.
9. Clean, descriptive product titles
"Osmia No. 4" tells an AI nothing. "Osmia No. 4 Cedar & Clay Face Soap" gives it category, material and use in one line. Keep internal SKU codes out of titles.
10. Visible contact information and policies
Recommendation systems favor stores that look accountable: a contact page, shipping and returns policies, a physical trace. Shopify generates the policy pages; make sure they're linked in your footer.
11. Consistent niche identity
A store that sells coffee, plus phone cases, plus supplements is hard to classify. The tighter your catalog's identity, the easier it is for an AI to know when you are the answer.
12. Reviews where they can be read
Review content on your product pages (with its own structured data, which most review apps emit) is third-party evidence an AI can cite. A great product with zero visible social proof competes one-handed.
Then measure, because guessing is the old way
The list above is the input. The output, whether ChatGPT actually recommends you, is checkable directly: run your store through the same scanner we used for the 82-store study. It asks the real buying questions for your catalog and shows you exactly where you stand.
Free, no install, no signup: foundrate.co/scan. Fix what it flags, wait a week, scan again.