Skip to content
GetAllAPIs
APIsPricingNetworkDocsPlaygroundFree tools
Dashboard Get started
APIsPricingNetworkDocsPlaygroundFree toolsDashboardGuidesFAQ Get started

Guides

  • Audit a website from an AI agent over MCP
  • How to read an audit report
  • Find competitor topic gaps from public sitemaps
  • robots.txt, noindex or canonical: which one to use
  • SEO checks before and after a site migration
  • Fix broken internal links and redirect chains
Home / Guides / How to read an audit report

How to read an audit report

Updated 2026-11-17

The report is designed to be acted on from the top down. You rarely need to read all of it.

Summary and scores

summary holds an overall score from 0 to 100, four sub-scores and a one-paragraph diagnosis. In the free site mode the weights are crawl and indexing 35%, on-page 30%, content 20% and performance 15%. Each finding subtracts points by severity, scaled by the share of sampled pages it affects, so one missing title on fifteen pages costs less than the same problem on every page.

Use the score to compare a site with itself over time, not with other sites. A 90 does not mean "ranks well"; it means few technical problems were found in the sample.

Actions and the plan

actions is the ranked to-do list. Each item has an impact, an effort, the number of pages affected, evidence URLs and the finding it resolves. Priority is roughly severity times reach divided by effort.

plan groups the same actions into now (high impact, or medium impact and low effort), next and later. If you only do one thing with a report, do the now list.

Severity

LevelMeaningExamples
highPages cannot be crawled, indexed or trusted as servedHomepage error, robots.txt blocking Googlebot, cross-domain canonical
mediumPages rank or convert worse than they shouldMissing descriptions, duplicate titles, thin pages, slow responses
lowWorth fixing when you are in the template anywayLong titles, missing HSTS, images without alt text

Every finding has its own page in the check reference with how we detect it and how to confirm the fix.

Evidence and sources

Each finding lists the URLs it was found on. sources lists every file or service the report used (pages, robots.txt, sitemaps, optional PageSpeed Insights) with the time it was fetched. If a client disputes a finding, the URL is the answer.

not_covered and partial

not_covered lists what the current data mode cannot measure, such as search volumes or backlinks in the free mode. We leave those fields empty rather than estimate them. partial: true means an optional source failed (for example PageSpeed timed out); the rest of the report is complete and the call is billed.

What to show a client

Diagnosis, the now list with evidence, and not_covered. Leave the raw site crawl block out unless they ask.

Run the full audit on your site. One call checks every item in the check reference and ranks the fixes.

Open the playground Endpoint details

Product

  • API catalog
  • Pricing
  • Site audit
  • Playground
  • Dashboard
  • Network

Free tools

  • All free tools
  • Title & description checker
  • robots.txt tester
  • Sitemap validator
  • hreflang checker
  • Heading outline
  • Structured data checker
  • Social preview
  • Indexability checker

Learn

  • Guides
  • Check reference
  • Use cases
  • Compare
  • Glossary

Developers

  • Docs
  • Quickstart
  • MCP server
  • Agent guide
  • llms.txt
  • OpenAPI

Company

  • About
  • FAQ
  • Changelog
  • Data sources
  • Contact
  • Security

Legal

  • Terms of Service
  • Privacy Policy
  • Refund Policy
  • Acceptable Use
  • Data Processing
GetAllAPIs © 2026 Mohamed Outerbah · [email protected]