What the crawler does
It walks your site the way a search engine would, up to three hundred pages, and runs sixty-eight checks on what it finds. Titles and meta descriptions, headings, canonical tags, redirects, broken links, images without alt text, indexing rules, sitemaps and structured data.
Nothing to install and nothing to configure. You give it the address and it reads what is actually published, not what your CMS thinks is published.
Four pillars, one score
- Crawlability — whether a machine can reach and index the page at all.
- Page clarity — whether the page states what it is about, in the places a parser looks.
- Quotability — whether there are passages an assistant can lift as an answer.
- Authorship — whether it is clear who wrote this and why they can be trusted.
Each is worth a quarter of the total, so you can see which part of the problem you have. A site can be perfectly crawlable and still unquotable, and that is a different week of work than a site nobody can index.
Why GEO is a separate job
Ranking in Google and being quoted by an assistant are not the same task. Google needs to reach the page, understand it and trust the domain. ChatGPT, Perplexity, Gemini and Claude need a passage short enough to lift, specific enough to be worth lifting, and attributable to someone.
So the GEO fixes are about structure and evidence: direct answers near the top, headings that match real questions, facts with numbers and sources, an author with a name.
Ranked fixes, not a wall of warnings
A technical audit usually produces hundreds of lines and no decision. Here the findings are ordered by what moves the score, with the pages they affect listed on each one.
Every fix comes with the change spelled out, so it can go straight to whoever builds — or into your AI builder with the rest of them in one prompt.
Run it again
Re-run the crawl after the changes ship. The score sits against the previous audit, with the line underneath saying what it gained or lost, so the work is measured rather than assumed.