The SEO Audit

A HubSEO audit is a measurement pass over one URL. It fetches the live page, renders it, runs Lighthouse lab tests, probes crawlability files, checks structured data and security headers, and measures AI-search readiness. The output is a prioritised, evidence-backed report: every finding names its source.

What gets measured

Collection. The engine runs these probes against your URL:

• An SSRF-pinned HTTP fetch of the page: the hostname is resolved by DNS first and private IP ranges are refused before any connection is made.
• Lighthouse lab runs in both mobile and desktop emulation.
• Text probes for robots.txt, sitemap.xml, and llms.txt.
• Captured screenshots.
• CrUX real-user field data (Core Web Vitals p75), when enough real Chrome users have visited the URL for a public field sample to exist.
• Google Search Console data, when you have connected a verified property.
• A SERP snapshot for context.

Detection. Check groups run over the collected evidence:

A sixth group runs the 11 AI-search signals that feed the AI readiness score.

Scoring and presentation. Each group is scored, findings are merged and deduplicated across the fix pack, SEO checks, and Lighthouse audits, then ranked by growth impact. The exact arithmetic is documented in How scores are computed.

How to read the report

The first screen is a human summary, not a scorecard: a plain-English headline such as "Your website is healthy overall. Three things deserve your attention first." Below it, findings are grouped into what to fix first, what is working well, and what can wait. Each finding follows the Evidence Desk structure.

Quotas and limits

When the anonymous allowance is reached, retries return a clear rate_limited outcome. That is a quota refusal, not a crawl failure. Subscription plans can also schedule recurring audits; see Plans.

Partial and failed audits

Honesty about failure is part of the product: