What is generative engine optimization?
Generative engine optimization (GEO) is the practice of optimizing your presence and content to appear in responses generated by AI-powered search systems. It’s also spelled generative engine optimisation in the UK and India, and it means the same thing.
The idea reached a wide audience through a 2023 research paper that introduced GEO as a framework, along with a benchmark called GEO-bench for testing it. A lot of advice has followed. Some of it holds up. Some of it doesn’t, and I’ll flag which is which.
How is GEO different from SEO?
SEO aims to rank pages in a list of results, while GEO aims to get your content included and cited inside an AI written answer. The two overlap a lot, and GEO builds on SEO rather than replacing it.
| SEO | GEO | |
|---|---|---|
| Goal | Higher rankings in search results | Inclusion in AI generated answers |
| Competes for | Position on a results page | A place in the AI generated output |
| Key tactics | Crawlability, keywords, search intent, backlinks | Clear, extractable writing, credible third party mentions, fresh content |
| Success metrics | Keyword rankings and organic traffic | AI visibility, mentions, citations and AI share of voice |
Table adapted from Semrush’s GEO vs SEO comparison. Conductor frames the output difference the same way: a list of links on one side, a synthesized, direct answer on the other.
What is the difference between GEO and AEO?
AEO means answer engine optimization and GEO means generative engine optimization, and Conductor says they deal with the same core concept. Conductor adds that strategies for one help with the other.
Pick one term for your own pages and use it the same way every time. I use GEO.
Which generative engine optimization strategies actually work?
The strategies with real support are original content with a point of view, quotes and statistics, clear page structure and a crawlable site. Most of it is still SEO, and Google says optimizing for generative AI search is still SEO.
I haven’t run a project under the banner of generative engine optimization. My work is brand-led: website design and development built around the brand, not GEO as a discipline. But every site I build already does what this guide describes: real content with a point of view, a clear heading hierarchy, fast and crawlable pages. That’s not GEO. It’s just the craft done properly, and it happens to be what GEO is asking for.
Publish original content with a point of view
Google asks for original, non-commodity content with a unique point of view, based on first-hand or expert experience. It also treats mass-producing pages to target query variations as a spam violation. One strong page beats ten thin ones.
Add quotes and statistics
In the research Semrush cites, pages containing quotes and statistics had 30 to 40% higher visibility in AI responses than pages without them. Only use a number you can link to, and link it right next to the claim.
Structure pages so a machine can follow them
Google recommends organizing pages clearly with headings and using semantic HTML where practical. This is where design and build meet. A clean heading hierarchy isn’t decoration. It’s the outline a machine reads.
Keep the site crawlable
Google says content should be publicly accessible and crawlable, and it points to JavaScript SEO practices. Semrush adds that server side rendering may matter, since AI crawlers may struggle with JavaScript. That’s one reason I care how a site is built, not only how it looks.
Keep content fresh and earn mentions
Semrush notes that updated content may be preferred by AI tools, and that brand mentions without links may still help AI visibility. Both are hedged in the source, so treat them as sensible habits, not guarantees.
Do I need schema markup or an llms.txt file for GEO?
No. Google says structured data isn’t required for generative AI search, that you don’t need special schema.org markup, and that you can ignore llms.txt and other AI specific files.
Google also says to ignore “chunking” content and rewriting pages just for AI. Structured data is still worth having as part of normal SEO, because it helps with rich result eligibility. Add it for that reason, not because an AI needs it.
How do I get cited in ChatGPT, Perplexity and Google AI Overviews?
No method guarantees a citation in any AI answer. For Google’s generative AI features your page needs to be indexed and eligible to show with a snippet, and even then indexing and serving aren’t guaranteed.
I can only back up the Google side with primary documentation. For ChatGPT and Perplexity I won’t give you steps I can’t source. What you can do is track results: Google offers a Generative AI performance report in Search Console. Semrush also suggests that presence on platforms like Reddit and YouTube may help, though it words that as a maybe too.
Is SEO dead now that AI answers exist?
No. Google says optimizing for generative AI search is still SEO, and Semrush says the two overlap considerably.
If your site is slow, hard to crawl or thin on real content, an AI engine has nothing good to pull from. Fix that first.
Are there statistics on GEO and AI citations?
The best known study found that GEO methods can boost visibility by up to 40% in generative engine responses. Its authors also say results vary by domain, so treat that figure as a ceiling from a benchmark, not a promise.
Be wary of any article that quotes a precise citation figure without linking the study. If you can’t open the source, don’t repeat the number.
Next step
If you want a site built to be read by people and by machines, tell me about your project. I design and build every site myself, from the first call to launch.
About the author. Satyam Dubey is the founder of The Design Shop, a one person design and development studio. He has about 10 years in design and has shipped 50+ products.