M

Method

Agent questions, not an SEO score.

The check asks the questions an AI agent asks: Can I read the page? Do I understand what is offered? Can I find a way to get in touch? A fast website with perfect meta tags will not score 95 if an agent cannot understand what it sells.

  1. Understandability 35 %

    Can an agent understand who you are, what you offer, who it is for, where you work, and what to do next?

  2. Discoverability 25 %

    Can search engines and AI search read your website? robots.txt, indexing, sitemap, and firewall.

  3. Operability 20 %

    Can an agent reach you? Contact methods, navigation, labeled forms, links, and buttons.

  4. Technical 10 %

    Is the content in the HTML? Headings, canonical URL, structured data, and language.

  5. Trust 10 %

    Legal notice, contact details, people, and references: who is behind the website?

Agent Ready means 90 points and no critical finding. One critical finding caps the score at 74, and two or more cap it at 59: an agent would get stuck there, however strong the rest is. Information an agent recognizes only vaguely counts for 30%, because it could not relay it confidently.

  • Homepage plus up to four subpages
  • JavaScript rendering
  • Analysis with a language model
  • Result reused for 24 hours