GEO and AI SEO expert: found, then cited
I help sites get cited by ChatGPT, Perplexity and Google AI Overviews without giving up the classic SEO that still drives most traffic. I build the diagnostic tools I use myself, and this site is my test ground: structured data, Markdown versions of pages, access for AI crawlers. In a project, I handle SEO at scoping, not after go-live.
Why online visibility is changing in nature
Google remains the main entry point for organic traffic for the vast majority of sites. But generative assistants (ChatGPT, Perplexity, Gemini) and the AI summaries in search results now answer some of the questions that used to generate a click. The risk is not that Google disappears, it is that part of informational traffic shifts to answers that do not always cite their source.
The answer is not two separate strategies but a common foundation: a clear structure, structured data, consistent internal linking and passages that can be cited out of context serve both at once. GEO adds requirements of its own, such as writing that answers the question first, but it never replaces the technical and editorial work of SEO. An llms.txt file remains optional: Google Search ignores it.
My method
I led this work myself with major accounts, alongside my delivery role. The method has four steps, from technical review to measurement.
Diagnose
Indexability, performance, access for AI crawlers, existing structured data.
Structure the content
Keywords and intent for SEO, passages that answer the question first for GEO.
Mark up the entity and the pages
Schema.org, an identified author, internal links that tie pages together.
Measure and iterate
Search Console, tracking of citations in generative engines, corrections.
In a web agency: SEO assisted by AI, never handed over to it
On a web project run in an agency, AI took its share of the SEO work: grouping queries by intent, suggesting page structures, preparing first drafts of tags and structured data. The gain is real on everything that repeats.
What stays human: choosing the topics worth covering, checking every fact, judging whether a passage really answers the question. Content produced without that review looks like everything else, and content that looks like everything else has little reason to be cited.
Tools I built for SEO and GEO
- SEO Diagnostic & ComparatorPageSpeed Insights measures performance. This diagnostic goes further: SEO tags, crawlability, GEO signals (citable passages for AI, optional llms.txt), content structure. Up to 3 pages analysed in parallel with side-by-side comparison.
- Site health scan: eco-design, accessibility, securityLoads a page in a real browser to measure its actual weight, estimated CO2 footprint, WCAG accessibility issues and missing security headers.
Frequently asked questions
What is the difference between SEO and GEO?
SEO (Search Engine Optimization) optimises content for classic search engines like Google. GEO (Generative Engine Optimization) adapts it to be cited by generative engines: ChatGPT, Perplexity, Google AI Overviews. Both coexist and call for a shared editorial strategy.
How long does it take to see SEO results?
Usually several months, depending on competition for the target queries and the site's starting point. SEO is a long-term investment: gains accumulate, unlike paid media, which stops with the budget.
Does an llms.txt file improve rankings?
Not for Google Search, which ignores it. It can help some assistants find a site's useful pages, so it is an optional extra, never a substitute for markup and content.