How GEO differs from SEO
SEO helps a page rank in regular Google results. GEO helps a piece of your text end up in the ready-made answer an AI generates. People may not click through, but they see your name as the source. One does not replace the other: without basic SEO, AI systems simply have nowhere to take your content from.
What actually works
- A short answer at the top of the page: 2-3 sentences that can be quoted without context.
- Specifics: numbers, timeframes, tool names and real examples instead of generic phrases.
- Structured data (schema.org): FAQ, services, organization, author.
- Question-and-answer blocks phrased the way people actually ask.
- Consistent business information across the site, social profiles and directories: the same name, services and contacts.
- Open access for search and AI crawlers in robots.txt, and indexing not only in Google but also in Bing, which powers ChatGPT search.
- Regular content for real search queries.
An llms.txt file is a useful extra with a short description of the site for AI systems. But it is not an official standard yet and not every system reads it, so do not rely on it alone.
A real example
For Kinezis, a rehabilitation equipment store, no ads were run. We built the technical foundation (FAQ markup, short answers, llms.txt) and publish content regularly - about 5 articles a week. Over 3 months, the Search Console AI features report showed 360 impressions of the site in Google generative answers.
How to measure it
- Search Console: the AI features report (impressions in Google generative answers).
- Google Analytics 4: referrals from chatgpt.com, perplexity.ai, copilot.microsoft.com.
- A manual check: ask an AI system a question your business answers and see whether it mentions you.