Documentation
Page Scan checks
Every Page Scan scores 157 positively weighted checks across four equally weighted categories. Page Experience and AI Discoverability inspect technical access and delivery; SEO and AI Content inspect readable meaning, intent, and AI interpretation.
Four check categories
Technical checks
Page Experience
25%Technical checks for browser delivery, performance, automated accessibility evidence, and Best Practices on Mobile and Desktop.
- Six Mobile and Desktop component checks plus 117 scored provider audits
- Performance, automated accessibility, and Best Practices evidence
- Mobile and Desktop evidence with fixed weights and explicit fallbacks
Content interpretation
SEO
25%Content-interpretation checks for readable meaning, audience, search intent, relevance, and agreement across visible and machine-readable representations.
- Search intent clarity and coverage
- Title, description, social, and structured representation fidelity
- Useful internal links, meaningful media, and terminology consistency
Technical checks
AI Discoverability
25%Technical checks for whether supported AI search and answer products can discover, retrieve, identify, and cite the page.
- Provider-specific robots access, robots.txt validity, and indexing controls
- Canonical identity, applicable hreflang, metadata, Schema.org, and Open Graph
- Main-content access, provenance, citation resolvability, and llms.txt orientation
Content interpretation
AI Content
25%Content-interpretation checks for whether an AI reader can understand the page, extract its answer, preserve qualifications, and act without inventing support.
- Comprehension, answerability, summary fidelity, and actionability
- Claim support, completeness, qualifications, integrity, and scope
- Observable credibility signals without declaring trustworthiness
Technical checks and interpretation checks
Page Experience and AI Discoverability answer technical questions: can browsers and supported machines retrieve, render, identify, and use the page? SEO and AI Content answer interpretation questions: is the page readable, is its purpose and intent clear, and can search or AI systems understand it without losing meaning or qualifications?
Evidence can support more than one explanation, but the versioned Ownership Registry gives every underlying defect one scored owner so the same problem cannot be deducted twice.
Current methodology versions
This overview describes Evaluation Version page-scan-2026-08-18-v16. Each category page links to its active immutable rubric. Historical Page Scans retain the exact Evaluation Version and rubric identities used when they ran, so a future methodology cannot rewrite a saved result.
- Page Experience
page-experience-rubric-v5- SEO
seo-rubric-v1- AI Discoverability
ai-discoverability-rubric-v3- AI Content
ai-content-rubric-v1
How to interpret the checks
These automated checks describe bounded evidence. They do not guarantee rankings, traffic, accessibility compliance, trustworthiness, inclusion, citation, recommendation, end-to-end agent completion, or any business outcome.
See how the four category scores combine on the Assessment Areas page.