39111906462099

The Technical Audit

The Technical Audit (formerly known as the Page Audit) is Gumshoe's tool for evaluating how well a page on your website is structured for AI-generated answers. As AI models like ChatGPT, Claude, Gemini, and Perplexity become go-to tools for discovery and decision-making, showing up in their answers matters just as much as ranking on Google.

Traditional SEO focused on keywords and backlinks to help pages rank in search results. The Technical Audit goes further: it evaluates whether AI models can find your content, understand it, and cite it when generating answers.

Why are there two versions of the Technical Audit?

Gumshoe runs two versions of the Technical Audit side by side, and every run gives you results from both. You'll find them on two tabs on the Analysis results page: Technical Audit (the original version, v1) and Technical Audit v2 Beta.

For the full comparison, see Why don't my Technical Audit and Technical Audit v2 scores match?.

What does the original Technical Audit (v1) measure?

When you run a Technical Audit on a URL, the original version evaluates it across six categories and produces an overall AI Optimization Score. Each category gets its own score along with a list of Highlights (what is working) and Recommendations (what to improve).

What does my AI Optimization Score mean?

Your AI Optimization Score reflects how well AI models can interpret and reference your page based on the six categories above. A lower score means there are improvements to make. In many cases, targeted changes to content structure or metadata can lead to meaningful gains without a full page rebuild.

The AI Optimization Score belongs to the original Technical Audit. Technical Audit v2 has its own score, built from its pass or fail checks, and the two aren't expected to match.

What are Technical Audit recommendations?

Technical Audit recommendations are the specific, actionable suggestions the original Technical Audit surfaces for each category after analyzing your page. Each recommendation identifies something that could make the page easier for AI models to crawl, understand, or cite.

In Technical Audit v2, the equivalent is each failed check. Every failed check explains why it matters for AI and how to fix it, and it passes once the fix is live, which makes v2 the clearer way to confirm a fix worked.

Working through recommendations and failed checks, then rerunning the Technical Audit after your changes are published, is the most direct way to improve your score over time. Starter and Build your own include 10 Technical Audit runs a month, and the Playbook (also known as the Pro plan) includes 50, counted in runs, not URLs. See What's included in the Starter, Build your own, and Playbook plans?.

What should I do next?

Pick the pages worth fixing first. How do I use my Technical Audit results to improve my pages? walks through which pages to start with and how to work through the results.