Soon customers won't ask "where do I find a good winter jacket" — they'll ask an AI agent to find and buy the best one for them. That means your store has to be readable and transactable by machines, not just by humans. This checklist shows how ready you are. Work through the points and mark what you're missing.
Checklist: 10 points for AI commerce readiness
- 1. Structured product feed — Do you have a machine-readable feed (CSV/TSV/XML/JSON) covering the whole catalog?
- 2. Unique identifiers — Does each product have a GTIN, UPC or MPN so the agent can match it against other sources?
- 3. Product schema (JSON-LD) — Do your product pages use Schema.org Product with price, availability and ratings?
- 4. Real-time price and stock — Is the feed refreshed frequently (ideally every 15 minutes) so the agent never recommends sold-out items?
- 5. Complete logistics data — Weight, dimensions, delivery regions, shipping cost and delivery time per product?
- 6. Variant handling — item_group_id, color, size and size system, so variants don't collide?
- 7. Platform and protocol integration — Are you connected to ChatGPT Shopping (ACP) and prepared for UCP via your platform?
- 8. Agent-friendly checkout — Can an agent complete a purchase without human-only steps like CAPTCHA or forced login?
- 9. Post-purchase webhooks — Do you expose order status, tracking and returns so the agent can follow up?
- 10. Trust signals and brand entity — Do you have reviews, ratings and a clear, consistent brand that AI recognizes as an authority?
How to read your score
0–3 points: Your store is effectively invisible to shopping agents. Start with the foundation — feed, identifiers and schema. 4–7 points: You're technically visible but losing on data quality or trust. 8–10 points: You're transactable — now AI visibility and brand authority decide whether you actually get chosen.
The point most people forget
Points 1–9 make you purchasable. Point 10 makes you chosen. Two products with identical feeds and prices are separated by reviews, sentiment and how clearly the brand reads as an authority to the AI model. That's where work on answer engine optimization and brand entity moves commercial numbers.
Next steps with CitationLab
Once the technical points are in place, CitationLab helps you measure what actually happens: which products AI recommends, how you stack up against competitors, and whether sentiment around your brand is positive. Also read what UCP is and how to get into ChatGPT Shopping.
Frequently asked questions
What is an AI shopping agent?
Where do I start if my store isn't ready?
Is being on Shopify enough?
Definitions used in this article
- AI shopping agent
- An AI shopping agent is an AI that discovers, compares and buys products on a user's behalf — for example in ChatGPT, Google AI Mode or Gemini — by reading machine-readable product data and transacting via standards such as UCP or ACP.
- Universal Commerce Protocol (UCP)
- The Universal Commerce Protocol (UCP) is the open, Google-led standard for agentic commerce that lets AI agents discover and buy products across stores through a single integration.
- Agentic Commerce Protocol (ACP)
- The Agentic Commerce Protocol (ACP) is the standard from OpenAI and Stripe that powers Instant Checkout in ChatGPT Shopping.
- Product schema
- Product schema is Schema.org structured data (usually in JSON-LD) on product pages that exposes price, availability and ratings in a machine-readable form so AI agents can match and evaluate the product.
- Brand entity
- A brand entity is the clear, consistent representation of a brand that AI models recognize as an authority, with defined attributes and relationships, separate from any single keyword.