AI Catalog Readiness scores how legible your product data is to the AI shopping channels Shopify now syndicates to, shows what is holding each product back, and fixes the weak spots in bulk.
Since Shopify switched Agentic Storefronts on by default in March 2026, eligible stores are discoverable inside ChatGPT, Microsoft Copilot, Google AI Mode and the Gemini app through Shopify Catalog, with nothing to install. What decides whether your products get matched to a shopper’s question is the data behind them: how specific the title is, how much the description actually says, which attributes are filled in, image coverage, and whether your metafields are mapped.
What It Checks
A score per product and a store level view of what is dragging it down, so the work can be prioritised instead of guessed at.
Product Legibility
Title specificity, description depth, attribute completeness and image coverage, scored per product with the reason for the score attached.
Attributes and Metafields
Which structured fields are populated, which are empty, and which custom metafields are not mapped into Shopify Catalog.
Channel and Policy Checks
Agentic Storefront settings, product disapprovals on connected channels, domain verification and store policy pages.
Where the Gaps Usually Are
Vague Titles
Titles written for a human browsing a collection page rarely carry the attributes an AI channel matches a question against.
Thin Descriptions
A single marketing line gives a shopping agent almost nothing to work with when a buyer asks for something specific.
Missing Attributes
Size, material, colour, fit, compatibility and use case, empty on most catalogues and exactly what comparison questions are built on.
Image Coverage
Products with one low quality image or no alternate angles, flagged so the shoot list writes itself.
Unmapped Metafields
Headless and custom builds often keep fit or material in metafields that Catalog cannot parse without explicit mapping.
Multi-Market Data
Catalog works from your default market’s product data, which catches stores selling across several countries off guard.
Bulk Fixing, With Your Own API Key
The audit is only half of it. Rewriting a catalogue by hand is what stops merchants acting on this kind of report, so the fixing is built in.
- Bulk rewrite of titles, descriptions and attributes, filtered by score, collection or vendor
- You bring your own AI provider key, so you pay model costs directly and there is no per word markup
- Everything is staged for review before it publishes, with a revert path per batch
- Planned after launch: attribution reporting for AI referred orders, which Shopify now tags in the admin
Who It Is For
Merchants who were opted into a new discovery channel without asking for it and have no way to tell whether it is working for them.
- Catalogues over a few hundred SKUs, where manual review is not realistic
- Stores with imported or supplier supplied product data of uneven quality
- Headless and custom builds keeping product attributes in metafields
- Merchants selling into several markets from one catalogue
Status and Honest Limits
Where it stands
- In development, no App Store listing yet
- Embedded admin app on the Admin GraphQL API
- Merchant supplied AI provider key, no model markup
- Billing through Shopify, monthly, in USD
What this is not
Pricing
Planned launch pricing, billed monthly through Shopify and banded by catalogue size. The audit and score are free, permanently, because a readiness score you have to pay for before seeing is worth nothing.
Audit
$0
- Full catalogue score
- Store level gap report
- No bulk fixing
Starter
$29
- Up to 500 products
- Bulk rewrite with your API key
- Staged review and revert
Growth
$79
- Up to 5,000 products
- Metafield mapping checks
- Channel and policy checks
Scale
$199
- Unlimited products
- Multi-market checks
- Roadmap features as they ship
- Priority support
Questions About This App
What are Agentic Storefronts?
Do I need to do anything to appear in ChatGPT?
Can people buy inside the chat?
Will rewriting my product data hurt my SEO?
Whose AI model does the rewriting?
When will it be available?
Find Out How Your Catalogue Scores
Tell us roughly how many products you sell, whether your store is headless, and which markets you sell into. The audit is where this starts.