E-mail:
jsem@patrikpilous.cz
My vision
GEO, or SEO for AI
GEO stands for Generative Engine Optimization. Put simply, it is SEO for LLMs, meaning AI models and answer engines. GEO is the process of making sure your digital content, website or brand gets maximum reach and visibility within generative AI models such as ChatGPT, Claude, Gemini, Perplexity and others. It is far from just a game with keywords. It is about semantic understanding, brand associations and working with entities. If someone asks for a tip on the best sauna in Prague or the best law firm, LLMs will put together a ranking for them. How do you get into it? How do you benefit from the ever-growing number of searches in ChatGPT and other tools? That is GEO. At its core, it does not differ from the principles of SEO and traditional marketing. Only the playing field is changing.
- Your brand is crucial for GEO. AI models form their picture of it from context and how often it is mentioned.
- Creating content for GEO is slightly different from creating it for SEO, but on the essential questions the right approach benefits both SEO and GEO. They certainly do not exclude each other, nor are they fundamentally different, as some people rather sensationally claim.
- Setting up associations and positioning is very important for GEO, more important than it used to be for SEO.
- The core principles still apply, so an experienced SEO consultant has a much better chance of getting you into AI models than an AI specialist.
Client testimonials
The difference between SEO and GEO
SEO is a tactic that has been widely recognised for decades. It focuses on improving websites (above all) to achieve higher rankings and, often, higher revenue thanks to good positions in organic search (SERP). This optimisation technique includes, among other things, keyword research, building the website architecture, creating top-quality content and building backlinks. SEO primarily revolves around the criteria that search engines such as Google and Seznam (Czech search engine) use to evaluate, categorise and index websites.
Over its lifetime, SEO has gone through a dramatic evolution. At the start, it was mostly about hacks that SEO consultants used to try to trick search engines. Today, SEO is more about mapping websites, meeting users’ needs as fully as possible and search intent.
GEO, on the other hand, is a newer concept that emerged in response to the rise of AI and AI-based answer engines, also known as generative search engines. These search engines use artificial intelligence algorithms to generate answers and resolve queries, rather than simply cataloguing and ranking websites based on keywords. GEO is essentially an optimisation tactic focused on creating content that not only engages human readers but also appeals to AI-driven search engines.
Unlike search engines (Google), answer engines (Perplexity) usually satisfy user intent within their own interfaces, but most of them also cite sources directly with links to the pages. Traffic will therefore certainly be lower than from search, but (for now) results in LLMs do not have to compete with paid placements.
How to do GEO?
Optimising for AI models involves far more work with your brand and PR than traditional SEO did. You need to gain the trust of LLMs so that they include you as a source when building an answer or mention you in their answers. Outputs in the style of modern PR and link building help greatly in breaking through the gate into the world of AI models and earning their trust.
Traditional search engines care much less about whether people write about you across the entire internet and focus more on the on-page side. This is where I see the fundamental difference.
At present, AI models rely on authority, the connection between a brand and a topic (associations), unique keywords on the page, technical terms, added statistics, surveys, cited sources and, of course, quotations. The most effective techniques include:
- Adding quotations
- Adding statistics
- Cited sources
- Technical terms
- Association with your field and topic
- Authoritative communication
The key to implementing GEO successfully is understanding how generative AI works and using that knowledge to create content that resonates with both people and machines. Here are a few tactics for effectively incorporating GEO into your digital marketing strategy
What customers ask about: Nothing differs from SEO here. Simply try to cover the full range of relevant and searched-for keywords.
Write opinions and share those of others: Your texts should express more of your own views and ideas. You should also quote others, share data and provide information with added value.
Distributing your pages and brand: LLM training is not limited to content published in posts on individual websites, but also covers communities such as Reddit. Distribute your content through these channels to maximise your ability to influence the narratives LLMs create in response to relevant questions.
The brand is the new link: The old practice of inserting keywords and links into pages has become somewhat outdated. A brand with an associated context is the new link. Even as a link builder, I sometimes aim only for a brand mention without a link. Yes, this is where the way things work has changed. Neither search engines nor AI models can be hacked by mindlessly scattering links. You need a more sophisticated approach.
Videos: Liven up your content by including visual and interactive elements such as videos, infographics and quizzes to engage both readers and AI-driven search engines. AI models also recognise when you are popular on social media and are more likely to show those brands.
If you feel that you are not appearing in AI models and need help with GEO or SEO, do not hesitate to get in touch.
Get in touch
