Skip to main contentSEO for AI Agents.
References

Every check traces to a primary source.

We don't grade on opinion. Every weight, threshold, and recommendation in the audit cites back to documentation published by Google, Microsoft, OpenAI, Anthropic, Perplexity, the W3C, or IETF. Here are the sources we use.

01 · Search engines + AI Overviews

Primary sources for how Google, Bing, and AI search surfaces discover, render, and cite the web.

  1. Google Search Central - SEO Starter GuideGoogle

    Baseline of Google's own SEO guidance. Reference for crawl, render, and indexing fundamentals.

    https://developers.google.com/search/docs/fundamentals/seo-starter-guide
  2. Google - About AI Overviews and AI ModeGoogle

    Official documentation on how AI Overviews are generated and how creators can influence inclusion.

    https://developers.google.com/search/docs/appearance/ai-features
  3. Google Search Quality Rater GuidelinesGoogle

    E-E-A-T definitions and rater rubric. The source-of-truth for what 'experience' and 'trust' mean to Google.

    https://services.google.com/fh/files/misc/hsw-sqrg.pdf
  4. Bing Webmaster GuidelinesMicrosoft

    Bing's crawl and ranking guidance. Bing powers ChatGPT Search's index, so this matters for AI citation.

    https://www.bing.com/webmasters/help/webmasters-guidelines-30fba23a
02 · AI crawler + citation protocols

How AI assistants identify themselves, request content, and decide what to cite.

  1. OpenAI - Overview of OpenAI CrawlersOpenAI

    OAI-SearchBot, ChatGPT-User, and GPTBot - identifiers, IP ranges, and opt-out behavior.

    https://platform.openai.com/docs/bots
  2. Anthropic - ClaudeBot crawlerAnthropic

    Claude's crawler identifier, behavior, and robots.txt expectations.

    https://docs.anthropic.com/en/docs/agents-and-tools/web-search-tool
  3. Perplexity - PerplexityBot and Perplexity-UserPerplexity

    Perplexity's two crawler identities and how each is used at index vs answer time.

    https://docs.perplexity.ai/guides/bots
  4. llms.txt proposalllmstxt.org

    Proposed standard for a site-level file telling LLMs what's safe to read and cite. Adopted by major AI tools.

    https://llmstxt.org/
03 · Structured data + schema

Specs we validate against when scoring schema density and richness.

  1. Schema.org full vocabularySchema.org

    Canonical vocabulary for structured data. Source-of-truth for every @type we accept.

    https://schema.org/docs/full.html
  2. Google - Structured data general guidelinesGoogle

    Required vs recommended properties for rich-result eligibility. Drives the AIO and Foundational check weights.

    https://developers.google.com/search/docs/appearance/structured-data/sd-policies
  3. JSON-LD 1.1 specificationW3C

    Wire format for the structured data we parse on every page.

    https://www.w3.org/TR/json-ld11/
04 · Performance + Core Web Vitals

Specs for the performance signals that feed our Technical dimension.

  1. web.dev - Core Web VitalsGoogle / Chrome team

    LCP, CLS, INP thresholds and measurement guidance. Our Technical dimension thresholds match.

    https://web.dev/articles/vitals
  2. PageSpeed Insights APIGoogle

    Lab + field data source we query during every audit when reachable.

    https://developers.google.com/speed/docs/insights/v5/get-started
  3. Chrome UX Report (CrUX)Google

    Field-data origin we cache to avoid re-querying PSI on every audit.

    https://developer.chrome.com/docs/crux
05 · Crawl + render

How modern engines actually fetch and execute the page.

  1. robots.txt specification (RFC 9309)IETF

    Standardised parsing rules. Our discovery layer respects this exactly.

    https://www.rfc-editor.org/rfc/rfc9309.html
  2. Google - JavaScript SEO basicsGoogle

    Two-pass indexing model. Justifies our render-vs-HTML diff check.

    https://developers.google.com/search/docs/crawling-indexing/javascript/javascript-seo-basics
  3. Sitemaps XML formatsitemaps.org

    Spec our sitemap generator and validator conform to.

    https://www.sitemaps.org/protocol.html

See how each source maps to a check

The methodology page documents which references inform which check and dimension, with the weight chosen for each.