What is a website audit and what should it include?

The page indexing report answers which URLs Google has and why it refused the rest. Core Web Vitals field data answers whether real visits clear the thresholds. A full crawl answers what exists, what 404s and what redirects twice. The internal link graph answers which pages nothing links to.

The first is the one most audits skip, and it is the only one that distinguishes not indexed from not ranking. Google's page indexing report documentation lists the exact reasons a URL is excluded — crawled but not indexed, duplicate without canonical, redirect, blocked by robots. A page Google never indexed cannot have a ranking problem.

The second has published numeric thresholds rather than an opinion: LCP 2.5s or less, INP 200ms or less, CLS 0.1 or less, at the 75th percentile of real loads. Field data, which is why a perfect lab score and a failing report are not a contradiction.

Three tests for the document you are handed. Every finding names a URL — "thin content" is not a finding, "these eleven URLs" is. Every claim carries its source; a severity rating with no measurement behind it is a preference. And the raw data comes with it: exports, not screenshots, because a screenshot cannot be re-run.

What an audit cannot tell you is why traffic fell, on its own. Google publishes a separate process for that, and it starts by separating a reporting change from a ranking change from a technical change. An audit that concludes you need a rebuild without that separation has skipped the diagnosis.