How to Pick GEO Experts: The Technical Camp, the Strategic Camp, and the Bridge Between Them

There's a concept that's been blowing up in marketing circles lately.

GEO. Generative Engine Optimization.

What does that mean?

Think about it. In the past, when you searched for something, Google gave you a list of links, and you clicked through them one by one. Now? You ask ChatGPT, you ask Perplexity, you ask Google's AI Overviews, and it hands you a paragraph — a single answer.

So here's the problem. If users stop clicking links altogether, how does your content get seen?

The answer is: get the AI to cite you directly. Make the AI see you as an authoritative source, and weave your information into its response.

That's GEO. Its goal isn't to rank on the first page of search results — it's to get "nominated" by the AI model.

This is a brand-new field. How new? So new that the people in the world who truly understand it, you can count them on your fingers — barely a dozen.

I've spent a lot of time lately digging through the work of the people at the very front of this wave. Today I want to talk about what they're actually doing, and what you can learn from them.

First, Two Camps

After reading through the work of these dozen or so experts, I noticed something really interesting.

They fall into two camps.

One is the technical camp. Michael King, David Konitzny, Dan Petrovic. What these folks do, in plain terms, is dissect things. They crack open Google's algorithms, dig out ChatGPT's internal prompts, and run experiments with data. The core question they care about is: how do AI systems actually work? Figure that out, and you'll know how to get your content chosen by them.

The other is the strategic camp. Duane Forrester, Marie Haynes, Gianluca Fiorelli. These people don't look at code — they look at trends. What they're thinking about is: when users stop clicking links, how does a brand survive? SEO turns from a tactical job into a strategic one. How should a company pivot?

You might ask, which camp is right?

Both are. But neither is enough on its own.

If you only understand the tech and not the strategy, you do the work and nobody knows what you're doing. If you only talk strategy and don't understand the tech, you sound convincing but nothing ever lands.

So the smartest move in GEO is to put both camps on the same team.

The Technical Camp: Cracking the AI Open for You

Let me tell you about three people.

Michael King, founder of iPullRank in the US. He coined a concept called "Relevance Engineering."

What is Relevance Engineering?

King's point is that from now on, content optimization can't just stare at a single Google search box. User search behavior has splintered. The search bar on TikTok, the chat box on ChatGPT, the question box on Perplexity — these are all "search interfaces." What you need to do is make the AI consider you authoritative in all of these places.

He even wrote AI Search Manual to systematize this framework. My takeaway after reading it: this person is turning something fuzzy into an actionable methodology.

David Konitzny, head of product at Kosch Klink Performance in Germany. His approach takes the road less traveled.

He's spent a huge amount of time doing one thing: reverse engineering.

What does that mean? He cracks open the systems behind Google and ChatGPT and analyzes how they process information internally. In other words, while everyone else is guessing how AI works, he's taking apart the actual code.

The value of this approach: he's not working from speculation, he's working from evidence.

Dan Petrovic, founder of DejanSEO in Australia. His method is experimentation. Lots of experimentation. Machine learning, natural language processing, algorithm testing — he works like a scientist, using data to understand how search engines actually operate.

These three share one thing: they believe the mechanisms behind AI search systems can be understood and cracked. Figure out the mechanism, and you can design the right strategy.

The Strategic Camp: A Different Angle on the Problem

Now let's talk about the other camp.

Marie Haynes, an SEO consultant in the US. She may be one of the people in the world who understands Google's algorithm updates best.

She wrote a book called SEO in the Gemini Era. The core idea: Google's Navboost system — the one that analyzes user clicks and interaction behavior — is one of the most important drivers of ranking.

Her advice is actually incredibly simple: build a website that users feel satisfied with after they click.

It sounds obvious, but a lot of people have lost sight of this fundamental truth. Everyone's busy studying algorithms and keywords, forgetting that the ultimate goal of a search engine is to serve user satisfaction.

Duane Forrester, who used to work on search products at Microsoft/Bing and now runs UnboundAnswers. His perspective is especially distinctive.

He said something that stuck with me: the future of marketing will shift from demographics to semantics.

What does that mean?

In the past, when you ran ads, you targeted by age, gender, location. In the future? You target by semantics. You reach people based on the concepts and questions they actually care about. Because the way AI understands the world is at the semantic level — it has nothing to do with demographic categories.

Gianluca Fiorelli, an SEO consultant in Italy. He coined a term: "Zero-Click SEO."

Users get the answer directly from the AI and never click through to your site. So does your content still have value?

Fiorelli says yes.

Even if nobody clicks, your content has to be cited by the AI, because the citation itself is brand exposure. You want the AI to mention your name, your viewpoint, your data when it answers users. Put bluntly: stop chasing traffic, chase brand.

The People Bridging Two Worlds

There are some experts who are hard to slot neatly into either camp. Their value lies precisely in "bridging."

Andrea Volpini, founder of WordLift in Italy. He studies knowledge graphs.

What's a knowledge graph? You can think of it as an enormous relationship network. A knowledge graph structures the connections between pieces of information, so that AI doesn't just "read" your content but genuinely "understands" what your content is about and how it relates to other things.

Volpini's belief: the future of search is built on knowledge graphs. Optimizing your content, put simply, means optimizing it for data structures, so the AI can give accurate answers based on verifiable facts.

Olaf Kopp, co-founder of Aufgesang in Germany. He did something incredibly valuable. He built a database of search engine patents and research papers.

Why does this matter?

Because the SEO industry is awash in claims — some backed by evidence, some just guesses. With a database of patents and papers, you can separate the trends that have real technical support from the ones that are just noise. Kopp was also among the first to study LLM optimization, publishing articles on it as early as 2023.

Malte Landwehr, CMO at Peec AI. He previously served as VP of SEO at Idealo and VP of Product at Searchmetrics. His signature trait is "scale thinking."

He wrote an article titled How to Optimize Content for LLMs and Get AI Citations. The core question he focuses on: how do large companies scale their GEO strategy? Not one page — a hundred thousand pages.

Don't Forget the Tool Builders

There's another group worth mentioning: the experts who build their own tools.

Mark Williams-Cook, based in the UK, partner at the agency Candour and founder of the tool AlsoAsked. His approach is pragmatic. Use AI to improve existing SEO workflows. Things like using AI to auto-generate redirect mappings and speed up site migrations. He's not reinventing SEO — he's making existing processes run faster.

Hanns Kronenberg, head of SEO at chefkoch.de. He's done a fair amount of research and publishing at the intersection of information retrieval and new search technologies.

These people remind us of one thing: GEO isn't pie in the sky. It has tools, it has methods, it has paths to execution.

Why Does Background Matter So Much?

After reading through all these experts' material, I came away with one feeling.

The field a person enters GEO from determines what they see.

Duane Forrester worked at Microsoft and contributed to Schema.org. So when he looks at search, he looks at it from the level of system infrastructure. What he cares about isn't "how to do it" but "where the whole system is heading."

Michael King calls himself "deeply technical and highly creative." So he can both crack open algorithms and translate technology into business strategy.

Fiorelli and Kopp run consultancies and agencies, so their perspective naturally leans toward client needs. What they produce tends to be more practical, more grounded.

Landwehr and Williams-Cook build products and tools, so they instinctively think "how do we scale this." What they produce tends to be more systematic, more replicable.

What does this mean? When you're looking for a GEO expert or learning GEO, you need to check whether their background matches your needs. You're a large company looking to scale? Find someone with a product mindset. You're an agency looking for new methods? Find someone from the hands-on camp.

GEO Isn't Just an American Game

One last point.

This list includes Americans, Germans, Italians, Australians, Brits. GEO isn't a Silicon Valley exclusive.

It actually makes perfect sense. AI search is global. ChatGPT is available worldwide, and Google AI Overviews has launched in multiple languages. No matter what market you're in, you can't sidestep this question: can your content be understood and cited by AI search systems?

For anyone doing content or building a brand, this is a test you can't avoid.

I've been thinking about this a lot lately. Maybe, looking back three years from now, the people paying attention to GEO today will turn out to be the same ones who first studied SEO all those years ago.

The early movers have one advantage: while nobody else is paying attention, they've already started stockpiling.

Stockpiling patents, stockpiling experimental data, stockpiling methodology. And the only thing that can buy these things is time.