How it works

One audit, from crawl to client-ready report.

SiteAuditorPro records what each worker actually measured and keeps the site-health analysis separate from the AI evidence layer.

Crawl and parse

We fetch the page, parse its HTML, and gather network-level technical signals (robots, sitemap, HTTPS, redirects, canonical).

We use PageSpeed Insights or an official Lighthouse mobile lab run. Real-user field data appears only when Google returns it.

Free audits generate a deterministic demonstration. Reviewed pilots record live OpenAI GPT and Claude answers, sources, and timestamps.

Findings roll up into two scores and a prioritized, client-ready report with evidence behind every line.

Methodology principles

Two scores, never blended

Search Health and AI Mention Share are reported separately. There is no single black-box visibility number.

Strict mention detection

Ambiguous AI mentions are excluded from strict score denominators and shown as needs-review evidence, not counted as wins.

Evidence mode disclosed

Every report states whether its AI layer is live, unavailable, or demonstration data. A demo is never presented as a live answer.

Scores describe the evidence collected in each dated audit. SiteAuditorPro does not present them as market-wide benchmarks or guaranteed outcomes.

Run a free audit and read the evidence behind every score.