Original research · June 2026
2,704 Shopify stores are AI-ready on paper. Most still lose the sale.
We scored 2,704 live stores the way ChatGPT, Claude, and Perplexity do. The median hit 89/100, a number that sounds finished. It isn't. It's propped up by the web basics Shopify ships for free, while the structured data an agent needs to actually answer a shopper and recommend a product is missing almost everywhere.
When a shopper asks an AI assistant to find, compare, or vet a product, the assistant answers from structured data, not from your design or your copy. A store that hasn't published that data simply isn't in the answer. We scored 2,704 live stores on the same free checker anyone can run, to see who's actually in the game.
Why the last 10% is the whole game
Your shopper is asking AI. Can your store answer?
AI assistants answer shopping questions straight from structured data. When it's missing, the agent can't speak for you, so it recommends the competitor who can. Here's how often these stores can actually answer, through an AI agent.
“How fast will it ship?”
Missing it? The agent can't quote delivery, so it recommends a store that can.
Powered by Shipping schema
“Can I return it if it doesn't fit?”
Missing it? No return terms to cite, so the agent hedges, or skips you.
Powered by Return-policy schema
“Is this actually any good?”
Missing it? Your reviews exist for humans but are invisible to AI, so you look unrated.
Powered by AggregateRating
“I've got a question about this product…”
Missing it? Nothing to answer from, so the agent moves on to a store that can.
Powered by FAQ schema
“Find me the exact one I mean.”
Missing it? No identifiers, so the agent can't match you to the product in question.
Powered by GTIN / MPN identifiers
“What does it cost right now?”
Missing it? No machine-readable price, so the agent can't put you in the comparison.
Powered by Offer price + availability
Median score across these stores is 89/100, which sounds finished. It isn't. That number is propped up by the web basics Shopify hands every store (titles, sitemaps, crawler access). The questions above are the part that decides whether an AI agent recommends you, and they're exactly the part most stores leave blank.
The score trap
A high score is mostly table stakes
Nearly every store nails the web basics Shopify ships by default (titles, sitemaps, crawler access), and that alone is worth a B. So the grade looks reassuring. The A-stores are the few that also added what AI actually reads; everyone else is coasting on the easy layer.
The gap that costs sales
The AI-critical layer most stores skip
Share of stores not passing each signal. These are the exact data points an agent needs to answer a shopper and put you in the recommendation, and most stores leave them blank. Every one tagged below is something AgentReady fills automatically.
By industry
How each industry stacks up
Median AI-readiness by vertical. Pick an industry to see where it leads and the specific gaps holding it back, the same view your store gets in the free checker.
Health, Wellness & Supplements
175 stores in this industry
Biggest gaps here
Shopify vs the field
Shopify is a 22-point head start
We ran the identical check against stores on other platforms. Shopify hands you the foundation the rest are missing, but the answerable layer above is still on you.
Shopify stores posted a median of 89/100, a 22-point lead over the next platform (WooCommerce, 67). The structured-data foundation Shopify ships by default is a real head start for AI shopping visibility.
Median AI-readiness score by detected platform. Bars show platforms with a meaningful sample (n ≥ 12); every store scored on the identical engine.
The fix
Close the gap: what to fix, and what it wins you
Each fix turns a shopper question you can't currently answer into one you can. That's the difference between being the recommendation and being skipped. Order matters: there's no point publishing perfect schema if the crawler that reads it is blocked, so work top to bottom.
- 1
Let the AI search crawlers in.
A site-wide block on OAI-SearchBot, PerplexityBot, or Googlebot makes a store invisible to that engine no matter how clean its data is. Allow the search bots; you can still block training bots like GPTBot if you want.
- 2
Publish Product + Offer schema that matches your catalog.
Agents read price, availability, and product details from structured data. When on-page schema and your live catalog disagree, AI engines down-rank both, so keep them in sync.
- 3
Put your ratings and reviews into schema.
Most stores show stars to humans but expose no AggregateRating an agent can cite. Publishing it is the single highest-leverage trust fix.
- 4
Add identifiers, FAQs, and shipping details.
GTIN/MPN identifiers help agents match your exact products; FAQ and shipping schema let them answer “how fast does this ship?” directly. These are the most common gaps in the study.
How we did it
In June 2026 we compiled a broad sample of live storefronts from public brand directories and category roundups across 12 industries, then scanned each with our public AI-readiness checker, keeping the 2,704that auto-detected as live Shopify stores. For the platform comparison we ran the identical engine against stores on other platforms. The checker reads each store's homepage and product pages the way an AI agent would and scores it across five dimensions on the same math as the live tool. This is a large convenience sample, not a probability sample; we report aggregates only and name no individual store. AI product recommendations also draw heavily on off-site sources (reviews, roundups, forums), so an on-store score is the controllable foundation, not the whole picture. You can run the exact same check on any store yourself.
How does your store score?
Run the same free check on your store and see your score, your percentile against your industry, and the exact fixes. No sign-up.