Shopify’s Free AI Shopping Readiness Tool: What Does It Check?

What is the tool?

Shopify describes it as a product page audit for agentic commerce, checking the structured data AI agents read to answer shoppers' questions. You enter a store URL, with no login, and get:

• An overall score from 0 to 100.
• A score for each of five categories.
• An impact-effort matrix listing each failed check, the likely fix, and how much it would move the score.

Independent write-ups date its launch to late April 2026, at a URL that now redirects to the shopify.com address. It works on any public storefront, so you can also scan competitors.

What does it check?

The five categories, as described by independent analyses of the scanner. Shopify's marketing and the in-tool labels use slightly different names:

According to one traffic capture of the scanner, it requests about 45 URLs, including the homepage, robots.txt, /llms.txt, /a/llms, /.well-known/ucp, several sitemaps, /products.json, a sample collection and around ten product pages. Some checks only test that a URL exists; others look at content, such as whether a shipping policy states real delivery times.

Why do most stores fail the same checks?

Write-ups of scans across many stores report three common failures:

1. No llms.txt. Early write-ups found the file missing at both paths the scanner checks. Since May 2026, Shopify has served /llms.txt on every store automatically, mirroring its /agents.md file, so this failure should now be rare on Shopify.
2. Incomplete product JSON-LD. Product markup often lacks brand, SKU, shipping details, item condition or ratings. A common cause is review apps that add star ratings with JavaScript after the page loads. The scanner reads the server's HTML, so anything injected later counts as missing.
3. Thin policy pages. Shipping and returns pages that exist but do not state windows, costs or timelines.

Independent scans put most small independent stores somewhere between 35 and 65, with stores that have already invested in structured data scoring higher. Treat those bands as rough guides, since Shopify does not publish benchmarks.

What should complete product markup include?

The product schema checks are where most stores lose points, so it helps to know what "complete" looks like. A product page's JSON-LD should carry, in the HTML the server sends:

• Identity: name, description, image, brand, and SKU, plus a GTIN where the product has one.
• Offer: price, currency, availability and item condition, updated whenever they change.
• Shipping and returns: shipping details and a return policy on the offer, matching your policy pages.
• Ratings: aggregate rating and reviews, if you collect them, rendered on the server.
• Variants: each variant's price, availability and image, not only the default one.

Google's merchant listing documentation uses the same fields, so fixing them helps Google Shopping and AI Overviews as well as the scanner. To check, open a product page, view the page source rather than the inspector, and search for "Product". Anything missing from the source is missing for most crawlers too.

Which checks matter most for AI visibility?

Not all of these checks are equally important for being recommended by AI assistants:

• Server-rendered product data: high. AI crawlers and shopping systems read prices, availability, brand and ratings. If those only appear after JavaScript runs, many systems miss them.
• Concrete policies: high. Shoppers ask assistants about delivery times and returns. Clear policy pages give assistants something accurate to quote.
• Transaction readiness: high for agentic checkout. Guest checkout and accurate stock matter when agents complete purchases, as with Shopify's Agentic Storefronts and Google's Universal Commerce Protocol.
• The UCP endpoint: relevant for Google checkout. The /.well-known/ucp profile supports checkout on eligible listings in Google's AI Mode and Gemini. It affects buying through agents, not whether you are recommended.
• llms.txt: low. Shopify now serves one by default, and the scanner rewards it, but Google says it ignores llms.txt, and there is little evidence that other AI systems rely on it. Customise it only if you want to add a brand description, as covered in how to add an llms.txt file on Shopify.

What does it not check?

The scanner looks at your store's technical readiness. It does not test:

• Whether AI assistants actually mention or recommend you for shopping questions.
• Your product feed in Google Merchant Center or how Shopify Catalog presents your products.
• Reviews and mentions on other sites, which strongly influence AI recommendations.
• Content quality, such as whether product pages answer comparison and use-case questions.
• Variant accuracy, such as the right image on the right colour.

For that, pair it with a visibility check. Apps that measure AI traffic and conversions on Shopify are compared in which Shopify apps track AI visibility, traffic and conversions.

How should you use it?

1. Scan your store and two competitors. The comparison shows whether a gap is yours alone or common in your category.
2. Fix the high-impact, low-effort items first. Typical examples are adding brand and SKU to product markup, and rewriting policies with real numbers.
3. Move review markup server-side if your review app injects it with JavaScript. Check your app's settings or ask its developer.
4. Confirm /.well-known/ucp and /llms.txt respond by visiting them in a browser.
5. Rescan to confirm each fix registered.
6. Then measure visibility. Run 20 shopping questions for your products in ChatGPT, Google AI Mode and Perplexity each month, and note whether you appear and what is said.

What might this look like in practice?

Consider an illustrative candle shop, Wickmoor, that scored 41. The matrix put two items at the top: product markup missing brand, SKU and ratings, and a 60-word shipping page. Its review app added star ratings with JavaScript, so the developer switched on the app's server-side markup option. The owner rewrote shipping and returns pages with delivery windows by region and a 30-day return period. A rescan scored 63. Over the next two months, AI answers to "soy candles with free UK delivery" began quoting Wickmoor's delivery terms correctly. The owner treated that as the real result, not the score.

What do people ask about Shopify's readiness tool?

Is Shopify's AI readiness scanner free?

Yes. It is free, needs no login, and works on any public store URL.

Does a high score mean AI will recommend my products?

No. Shopify says the signals do not guarantee that AI agents will surface your products.

Does it fix problems automatically?

No. It diagnoses and prioritises. Fixes happen in your theme, apps, product data and policy pages.

Do I need llms.txt?

Shopify stores already serve one automatically, and the scanner rewards it. There is little evidence that it improves AI visibility on its own.