AIScan.site
Retail & ecommerce

AI readiness for online stores

A shopping agent compares three retailers in one turn. Two publish price; one does not.

M4C2C3D2D1B2

The moment

Tom, who runs a 400-SKU homeware store on his own

A buyer asks an assistant to find a specific stoneware dinner set under a certain price, in stock, delivered this week.

What the agent needs: Price, currency, availability and delivery, per product, without executing the theme.

What actually happens: Tom's price is rendered by a currency-converter script and his stock state is a JavaScript variable. The agent reads two competitors' Offer objects, cannot read his, and quietly drops him from the comparison. Nothing appears in his analytics, because the visit never happened.

Which dimensions decide it, and why

AIScan grades five dimensions. These are the ones that carry the outcome for this kind of business.

Commerce

This is the deciding dimension for a store. Check M4 asks whether price, currency and availability are machine-readable in the served HTML. Fail it and you are not expensive in an agent's comparison — you are absent from it.

Content

A catalogue with no llms.txt gives an agent no map. It will read a handful of URLs; you want those to be your collections, not your returns policy. Product descriptions also need to survive without the theme's JavaScript.

Discoverability

Product and collection URLs must be in the sitemap and reachable. Faceted URLs blocked wholesale in robots.txt frequently take real collection pages with them.

The checks that matter here

Fix guides: /docs/checks/commerce · /docs/platforms/shopify · /docs/checks/content

Most sites like this run on Shopify — read the Shopify playbook for where each fix lives.

Questions we get asked

Do AI shopping agents actually buy things yet?
Mostly they shortlist and hand off. That still decides the sale, because the shortlist is built from what the agent could read — and a product it cannot price is a product it cannot compare.
What exactly does check M4 want?
A price, a currency and an availability value that are readable from the HTML the server sends — normally a JSON-LD Offer on the product page. Prices painted in by a converter script do not count.
We hide prices behind a login for trade customers. Are we penalised?
For the pages that are genuinely gated, no — but publish a public price range or an indicative tier somewhere readable, or an agent has nothing to place you with.
Should our llms.txt list every product?
No. List collections, your key policy pages and the handful of products that define the range. It is a map, not a feed — the sitemap already carries the full list.
Do we need UCP or x402 to be agent-ready?
Not today. Both are bonus checks that can only add points. Get M4, C2 and C3 right first; they are what agents use now.
Our stock changes hourly. Won't structured data go stale?
Only if it is hardcoded. Render availability from the same variable the page uses and it stays accurate by construction.
How do we compare against competitors?
Scan them. npx aiscan-cli competitor.com works on any public URL and returns the same graded report you get for your own store.
Where do we start if we only have an hour?
Confirm the served HTML contains price, currency and availability on one product page. If it does not, that single fix moves more than everything else combined.

Scan your site and see where you actually stand

Free, no account, about twenty seconds. The report names every failing check by ID, shows the evidence we found, and gives the fix for your platform — plus a hand-off prompt you can paste straight into Claude Code or Cursor.

Run a free scan npx aiscan-cli yoursite.com

Prefer to read first? Browse every check we run or the guide library.

Next in Retail & ecommerceAI readiness for local shops with an online window

Also covers: online store, ecommerce, e-commerce, webshop, retailer, boutique, dtc, direct to consumer, product catalog, catalogue, dropshipping, print on demand, marketplace seller, homeware, fashion store, electronics store, grocery delivery, subscription box, shopping agent, agentic checkout.