Questions, answered
The things people actually ask us, answered the same way we score: plainly.
What is GEO, and why should I care?
GEO (Generative Engine Optimization) is making your business visible in AI answers: ChatGPT, Gemini, Google's AI Overviews. Buyers increasingly ask assistants what to buy and where; if the engines can't read or don't trust your site, they recommend someone else. GEO is SEO's next chapter, measured against answers instead of rankings.
What exactly does the free audit check?
Seven deterministic checks plus one AI content pass over at most 9 pages of your site: AI-crawler access in robots.txt, structured data (JSON-LD), freshness signals, semantic HTML, platform detection, sitemap discovery, page performance, and whether your content answers questions with checkable facts. Every finding comes with the evidence and a concrete fix.
My score is low. Is something broken?
Probably not. The score is honest by design. Common causes of a low score are real: robots.txt blocking major AI crawlers, no dateModified anywhere, missing FAQ/LocalBusiness/Article schema. The report shows the exact measured evidence behind every deduction, so you can verify each one yourself.
Will the audit crawl or overload my site?
No. It reads at most 9 pages (homepage, up to 3 categories, up to 5 products) with plain HTTP requests: no headless browser, no site-wide crawl. The load is comparable to one visitor clicking through a few pages.
Where does the monitoring data come from?
Real assistant interfaces. ChatGPT and Gemini answers are collected by scraping the live consumer apps (with web search) through a managed data provider, and Google AI Overviews come from real result pages. We archive every answer verbatim, so you can always read exactly what the AI said.
Why does my own ChatGPT show something different?
Personalization. Assistants adapt to your chat history, account and location, so no two people see identical answers. Our runs are collected from a neutral session located in your market, a fair baseline, but your personal view may still differ. That variance is inherent to AI answers, not a measurement error.
How often does monitoring update?
Runs are on-demand today, with a weekly schedule as the intended cadence. AI answers shift with the web, and weekly is frequent enough to catch real movement without chasing noise. Trends become meaningful after two or more runs.
I got my report. How do I actually fix things?
Start with the top priorities: each one explains the issue, the evidence, and what you gain by fixing it. Structured-data findings include ready-to-paste JSON-LD, validated for the exact schema type. Fix, re-run the audit to confirm, then watch your visibility trend in monitoring.
Email palma.alexandros@gmail.com. A founder reads every message, usually within a day. Include your audit URL or run date if it's about a specific result.
Looking for metric definitions? See the docs. Curious how the score works under the hood? Read the method post.