commercetools Agentic Commerce Readiness

Agentic commerce readiness check

Is your store ready for AI shopping agents?

Shoppers now ask agents to find, compare and buy for them. Check how well an agent can discover your store, read your products, evaluate your offers and complete a purchase.

Public storefronts only. A check sends about 30 requests and takes up to 30 seconds. It never signs in, adds to a cart or places an order.

Four questions an agent asks before it buys

The check follows a shopping agent's journey. Each stage has its own score, so you can see where agents get stuck.

  1. 1

    Discover

    Can agents reach the store and find its catalogue? We look at access rules for AI agents, sitemaps, and machine-readable guides such as llms.txt.

  2. 2

    Understand

    Can agents read a product without a browser? We look at server-rendered content, product structured data, identifiers and images.

  3. 3

    Evaluate

    Can agents compare the offer? We look at price, availability, shipping, returns and ratings in a form agents can trust.

  4. 4

    Transact

    Can agents act? We look for commerce protocols and interfaces: MCP servers, agent checkout profiles, catalogue APIs and feeds.

What the score means

The score is a snapshot of signals we can observe from outside, without logging in or placing an order. It is not a certification, a security review or a measure of how good your store is for people.

How to use it

Start with the fixes listed first: they unblock the most agent journeys. Emerging protocols count for less today and are flagged so you can plan for them.