The AI Optimization page is the original Technical Audit (v1) view of a page, on the Technical Audit tab of your Analysis results. It gives the page an AI Optimization (AIO) Score, a breakdown of six optimization categories, and Recommendations for making the page easier for AI models like ChatGPT, Gemini, Claude, and Perplexity to read. Your home page is scored by default, and you can run the Technical Audit on other URLs too.
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.
- The original Technical Audit (v1) is AI-graded feedback. An AI model reviews a structured extract of your page and writes Highlights and Recommendations across six categories. Because a model is making the call, results can vary a little between runs on a page that hasn't changed.
- Technical Audit v2 runs about 20 pass or fail checks. Each result shows the evidence Gumshoe found, a reason code, why it matters for AI, and how to fix it. The same page gets the same result.
- The two scores use different formulas, so they aren't expected to match or move together.
For the full comparison, see Why don't my Technical Audit and Technical Audit v2 scores match?.
What does the AI Optimization page measure?
- AI Optimization Score (AIO): your overall score, reflecting how AI-friendly the page is. A higher score means the page is well-structured and easier for AI models to interpret and reference in their answers.
- Structured data: whether the page provides structured data (like product or organization markup) that AI can parse. A low score may indicate missing or incomplete structured content.
- Page layout: the organization of headings, sections, and navigation elements. A clear, consistent layout helps AI models understand and follow the page's structure.
- Schema markup: how well the page uses schema to define and describe content. Good schema helps models extract product names, brands, categories, and page types.
- Navigation: the simplicity and clarity of your site navigation. Clean menus and link structures help AI models map your pages accurately.
- Content balance: how well you blend descriptive content with promotional material. Too much marketing copy and too little substance may reduce AI understanding.
- Metadata: whether you use proper meta titles and descriptions. These help AI models categorize your pages and associate them with relevant answers.
How do I use this?
- Use the Recommendations under each category to improve the page's AI readability.
- Open the Technical Audit v2 Beta tab for pass or fail results that show exactly what to fix.
- Run the Technical Audit again on the page after your changes are published to track improvements. Starter includes 10 Technical Audit runs a month and Pro includes 50, counted in runs, not URLs.
- Monitor how your brand appears in AI answers over time.
What should I do next?
Decide which fixes to make 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.