Agent Experience (AX) is how AI agents perceive, evaluate, and interact with your brand. It is the new UX — but for machines that buy, recommend, and decide on behalf of people.[S1]
For two decades, brands optimised for human attention: pages, ads, conversion funnels. That work is not going away, but a new audience has joined the queue. When a customer asks ChatGPT, Claude, or Perplexity for a recommendation, an agent reads your site, your schema, your reviews, and your documentation — then returns a single answer. You either appear in that answer, or you do not.[S2]
AX is measured on four dimensions — the same axes the Agent Readiness Index uses to score brands:
| Dimension | What it is | Test | Common gap |
|---|---|---|---|
| Discoverability | Can the agent find you? | Schema, llms.txt, sitemap | JS-only content |
| Completability | Can it complete the task? | Forms, APIs, structured data | Human-only flows |
| Actionability | Can it act on your behalf? | Booking, checkout, transact | Auth walls |
| Evidence | Why should it pick you? | Citations, freshness, proof | No third-party signal |
Discoverability rests on machine-readable surface area — schema.org, llms.txt,[S3] clean HTML. Completability and actionability extend the surface from “readable” to “operable”. Evidence — citations, dates, third-party mentions — is what tips a tie when the agent has to pick one answer.
Agent Experience is not a marketing layer bolted onto SEO. It is a product surface, and it sits inside a broader shift toward hybrid organisations where machines are first-class participants on both sides of the table. The same role-design logic that makes an internal agent useful — captured in the ABC Framework — is what makes an external agent recommend you. Both rely on legibility, ownership of an outcome, and evidence the agent can cite. If your brand is not legible to agents, you are invisible to a growing share of buyers, and that share is not coming back to manual search.
