Methodology

How we score AI visibility

Points are weighted by what independent testing shows drives AI answer pickup: retrievability and crawl access 25, answerable content quality 25, structured data 20, classical SEO hygiene 15, FAQ and Q&A coverage 10, experimental AI crawler conventions 5. Structured data is scored per page: each successfully scanned page earns zero (none), half (weak) or full (sound) credit, averaged equally. Free scans sample up to five pages; unknown pages are disclosed, not scored as missing. This is our assessment rubric, not a Google ranking formula, and does not guarantee discovery or citations. Saved reports retain their original rubric.

Crawlability — 15 diagnostic points

We fetch your robots.txt and check whether GPTBot, OAI-SearchBot, PerplexityBot, Google-Extended, ClaudeBot, and CCBot are allowed. Block any of them and a chunk of the AI ecosystem can't see you.

Classical SEO hygiene — 25 diagnostic points

Title length, meta description, single H1, canonical, OpenGraph trio, and image alt text. Boring fundamentals — and most sites still fail them.

Structured data — 20 of 100 points

We score JSON-LD in three states, not two. None means no markup. Weak means markup that needs improvement: it won't parse, it only calls the page a generic web page, or it leaves out a property Google requires for its type — half credit. Sound means it parses, uses a type that fits the page, and has the required properties. We name what's missing; we never guess a value to make a page pass.

AI crawler readiness — 15 diagnostic points

Content available in HTML helps crawlers that do not execute JavaScript (12 points). llms.txt (2 points) and ai:* tags (1 point) are optional experimental bonuses, not accepted standards or requirements. Their adoption and visibility benefits are not established. Supporting categories together contribute 40% of the overall score, proportional to their available diagnostic points.

FAQ & Q&A coverage — 15 diagnostic points

Visible questions, useful answers, and FAQ markup support answer extraction; their presence does not guarantee citation.

Content comprehension — 15 diagnostic points

We ask what the company does, who it serves, and what visitors should do next. Each criterion earns five diagnostic points only with an exact page quote and a separate AI support check. Unavailable evaluations are excluded, not marked as failed. This tests content comprehension, not discovery, trust, citations, or JSON-LD quality.

Bands

  • 80–100 Visible — Strong assessed readiness; requires sound markup on every scanned page.
  • 50–79 Partially visible — Readiness gaps remain, or markup needs improvement.
  • 0–49 Invisible — Significant gaps under this rubric; not proof of exclusion from AI answers.