Learn / For your team

Agent readiness for ecommerce managers

Much of agent readiness sits with whoever owns the catalogue rather than with engineering: published policies, complete product fields, accurate availability and a feed that covers everything you sell. The parts that do need engineering are few and specific, which makes them easy to ask for precisely rather than as a general concern.

A useful amount of agent readiness sits with whoever owns the catalogue rather than with engineering, which means an ecommerce manager can move most of it without waiting for a sprint.

What you own outright

  • Policies as text. Returns window, who pays return postage, shipping cost and threshold. See how to publish policies as text, not images.
  • Product fields. Every field your platform offers, filled. An empty field is a constraint you cannot answer.
  • Availability states. Backorder and preorder used accurately rather than collapsed into in stock.
  • Feed coverage. Item count against catalogue count, which is a five minute check that frequently finds a real gap.
  • Specifics in copy. Materials, dimensions, compatibility, care.

What you have to ask for

Four things, and they are specific enough to be tickets:

  1. Price and availability emitted as structured data in the served markup, at variant level.
  2. A variant picker that works without a pointer and without scripts attaching the options.
  3. An add to cart that is a real form rather than a click handler.
  4. Bot rules and CDN categories reviewed and set deliberately.

See how to brief your developer on agent readiness.

Check the bot rules first

This is the one to escalate this week regardless of everything else, because it is binary. If agents cannot fetch your pages, the quality of your catalogue is irrelevant, and the fix is a configuration change rather than a project.

How to keep it from decaying

Tie a short check to theme changes, app installs and checkout changes rather than to a calendar. That is what breaks it, and a monthly review finds the damage three weeks after anybody could still remember what caused it.

See how often should you test your store with AI agents.

Questions

What can I fix without a developer?

More than you would expect. Policies written as text, product fields filled in, availability states used accurately, feed coverage checked, and specifics added to descriptions. That is most of a first pass.

What do I have to escalate?

Structured data emitted in the served markup, the variant picker and add to cart behaviour, and the bot rules. Four items, each specific enough to be a ticket rather than a project.

How do I know if this is urgent for us?

Check the bot rules first. If agents are being blocked, everything else is theoretical and the fix is a configuration change rather than a roadmap item.

Scan your store

Agent readiness for ecommerce managers · Ottom