The five areas of agent readiness: find, read, act, pay, trust
A three-minute lesson on the framework behind the self-check.
Am I Ready for Agents? editors · 2 September 2026 · Reviewed September 2026
Agent readiness breaks into five questions: can agents find you, read you, act on your site, pay, and trust you. A business needs all five for an agent to finish a task, so the weakest area matters more than the average.
Why five areas?
Different organisations slice the subject differently. Cloudflare talks about a web that is readable, discoverable, callable and payable. ora.ai scores discovery, access, usability and payments. We use five plain questions so that a business leader can answer them without technical knowledge, and we add trust, because verification and public policies decide whether an agent's work is accepted.
What does each area ask?
- Find: are AI crawlers and assistants allowed in, and can they locate the pages that matter?
- Read: are products, prices, policies and articles available as text, without scripts or images standing in the way?
- Act: can the main task be done with standard forms, or through an interface built for agents?
- Pay: is the full price visible, and does your business accept a purchase an agent starts?
- Trust: can you tell verified agents from impostors, and are your terms public?
Why does the weakest area decide?
An agent's task is a chain. If it finds your hotel, reads the rates and fills in the dates, but your payment page blocks it, the booking fails and the customer's agent tries someone else. That is why the readiness levels rubric caps your level at one above your weakest area.
How do the five areas map to other frameworks?
Loosely, and on purpose. Find covers what Cloudflare calls discoverable and what ora.ai calls discovery. Read matches readable and access. Act covers callable and usability. Pay matches payable and payments. Trust has no single counterpart: it gathers verification, which Cloudflare's scanner checks under bot access control, and the public policies an agent and its user rely on.
The mapping is ours. Scanners and our rubric do not produce comparable numbers, so use the areas to organise questions, not to translate scores.
What should you do next?
Take the self-check, note your weakest area, and read the matching section of ask your web team.
If your team wants a quicker way to test scenarios, the readiness calculator uses the same five areas and shows which one caps your level.
Keep a note of your five percentages, because the self-check does not store them. The next lesson looks at the walls agents most often hit.