Does it check the whole site?

What one run fetches

A run is five fetches and a parse. The audit fetches the page you gave it as a desktop browser would, then four files from the site root: robots.txt, llms.txt, sitemap.xml, and the OKF bundle. Those four are site-wide, so a problem in robots.txt shows up on every page you audit, but the page-level checks, the headings, the body text, the internal links, the Open Graph tags, are scored for that one URL only. A run takes seconds, and every run is kept.

Why we audit pages and templates, not sites

Most sites are a handful of templates rendered many times. A product page, a service page, a location page, an article: fix the template and every page built on it changes with it. So rather than crawl a whole site and hand back thousands of rows that say the same thing, our team picks the pages that matter for your brand, audits them, and audits the templates behind them.

In Managed GEO those are your key pages, and we ship the fixes ourselves. In GEO Consulting and Trends, where a company may have many properties, we audit the templates and hand your engineers a prioritised list with the check names on it, so each fix can be verified in the same tool.

What falls outside a run

The audit reads what your robots.txt says rather than visiting as each crawler in turn, so a firewall that turns bots away by user agent will not show up in the crawlability check. If that firewall turns our auditor away as well, the run ends unreachable and says so. The audit does not discover pages on its own: you give it a URL, and it scores that URL.