AI assistants quote sources they can read and verify. Write factual sentences about yourself, back results with numbers, add Person, FAQPage and BlogPosting schema, and keep publishing.
Generative engine optimisation (GEO) means making your website easy for AI assistants like ChatGPT, Gemini, Perplexity and Google AI Overviews to understand, trust and quote. For a personal brand, it comes down to three things: say who you are in plain factual sentences, back every claim with structured data, and publish fresh, specific content regularly. I rebuilt my own profile site this way, and here is exactly what I did.
Why GEO matters for a personal brand now
Recruiters and founders no longer only Google a name. More and more, they ask an AI assistant: "Who are the strongest growth leaders in fintech in the GCC?" or "What has Vishal Gupta actually delivered?" The assistant answers from whatever it can read and verify. If your site is a beautiful image-heavy page with vague copy, the AI has nothing to quote, so it quotes someone else.
Classic SEO still matters, because AI engines lean heavily on pages that already rank. GEO sits on top of it. SEO gets you found; GEO gets you cited.
How do you make a personal website readable for AI search?
I treat the site as a source document, not a brochure. Every page has to answer a question an AI might be asked about me, in words it can lift directly.
1. Write one factual paragraph about yourself
Not a tagline. A third-person, encyclopedia-style paragraph: role, years of experience, industries, location, current employer and education. AI systems love sentences that read like a fact, such as "Vishal Gupta is a growth, go-to-market and marketing leader based in Muscat, Oman, with 11+ years across fintech, digital payments, agritech and D2C."
2. Put your results in numbers, with context
"Drove growth" is unquotable. "Cut customer acquisition cost from INR 5,000 to INR 650" is. Every result on my results page pairs a number with what moved and where.
3. Add structured data on every page
Schema.org markup tells machines what the page is about without guessing. The types that did the heavy lifting for me:
| Schema type | What it tells AI and search engines |
|---|---|
| Person | Name, job title, employer, location, education, LinkedIn, topics you know about |
| ProfilePage | This page is the canonical profile of that person |
| FAQPage | Ready-made question and answer pairs an assistant can quote |
| BlogPosting | Each article, its author, date and topic |
| BreadcrumbList | How the pages relate to each other |
The knowsAbout field on Person is underrated. It is the closest thing to telling an AI "these are my areas of expertise" in a format it trusts.
4. Answer the questions recruiters actually ask
I added a short FAQ: who I am, what I have delivered, which industries, what roles I am open to, how to reach me. Each answer is self-contained, so it still makes sense when an assistant quotes it out of context.
5. Add an llms.txt file
An llms.txt file at the root of the domain is a plain-text summary of who you are and which pages matter, written for language models. It is an emerging convention rather than a standard, but it costs ten minutes and removes all ambiguity.
The hosting mistake that quietly kills SEO
My first instinct was to keep the site on a drag-and-drop builder and paste custom code into an HTML embed. The problem: on most builders, embedded code runs inside an iframe, and search engines treat iframe content as belonging to a different document. Your carefully written copy and schema may simply not count.
I moved to static hosting instead, where every page is real HTML that crawlers read directly. Same design, far better visibility.
Fresh content is the part most people skip
A profile page is static by nature, and both search engines and AI assistants favour sources that keep publishing. So I set up a daily insights section, built from real problems I work on, with the confidential details stripped out. Each post is structured for GEO: a direct answer in the first paragraph, question-style headings, a comparison table or framework where it helps, and a short FAQ.
SEO gets you found. GEO gets you quoted. Consistent, specific publishing is what makes you the source.
A 30-minute GEO checklist for your own site
- Write a factual third-person paragraph about yourself and put it on the home page.
- List your top five results with numbers, context and company.
- Add Person and ProfilePage schema, including
sameAspointing to your LinkedIn. - Add a five-question FAQ with FAQPage schema.
- Publish an
llms.txtand a sitemap, then submit the sitemap in Google Search Console. - Make sure your pages are real HTML, not content trapped inside an iframe.
- Commit to publishing something specific at least weekly.
If you are hiring for growth and want to see how this thinking applies to a business rather than a personal brand, my AI toolkit covers how I use GEO and other AI capabilities in marketing work.
Questions
What is generative engine optimisation (GEO)?
GEO is the practice of making content easy for AI assistants such as ChatGPT, Gemini and Perplexity to understand, trust and quote in their answers. It builds on SEO with factual phrasing, structured data and question-led content.
Is GEO different from SEO?
They overlap. SEO helps your pages rank in search results; GEO helps AI systems extract and cite your content. AI engines still lean on well-ranked pages, so you need both.
Does llms.txt improve AI visibility?
llms.txt is an emerging convention, not a formal standard, so its impact varies by AI system. It is cheap to add and gives language models a clean summary of who you are and which pages matter.
Why does a website builder HTML embed hurt SEO?
Many builders place embedded code inside an iframe, which search engines treat as a separate document. Text and schema inside it may not be credited to your page.
Growth, GTM and marketing leader with 11+ years across fintech, payments, agritech and D2C. Based in Muscat, Oman.
See my case studies