GEO: 3 Major Opportunities for Greater Visibility in AI Responses

The era in which we received a lot of traffic from Google is over. So is the era of traditional search engines with the familiar 10 blue links. But how do we deal with this? How do we optimize websites when the search results page is increasingly being replaced by an AI-generated answer? And what does visibility even mean when users get the answer without clicking on a search result?

This is exactly where GEO comes into play. GEO stands for Generative Engine Optimization. Some call it LLMO, AI SEO, AEO, or Relevance Engineering. The name itself is secondary. What matters is how brands, products, and content become visible in generative AI systems. Because even though clicks from AI Overviews, ChatGPT, Gemini, or Perplexity currently do little to compensate for the lost Google traffic, new opportunities are emerging elsewhere.

While clicks from organic traffic are declining, we see 3 areas with particular potential for website operators:

One of the best SEO strategies of the past 10 years has been optimizing for informational search queries. Early in the decision-making process, users search for possible solutions, compare pros and cons, and gather information. This was exactly where companies could build visibility, establish trust, and later drive conversions with helpful content.

This research phase hasn’t disappeared—but it has shifted. More and more people are asking their initial questions not on Google, but through AI chatbots. To gather information, users no longer need to visit websites and read the content there. They get an initial overview directly from AI. They only turn to Google when it comes to comparing prices and making a purchase. 

At first glance, this is troubling for website operators because it results in fewer clicks. At the same time, many users arrive at the website significantly better informed when they do eventually click through.

We’re seeing this pattern more and more often in analytics data: Traffic is declining, but the conversion rate is rising. So fewer visitors don’t automatically mean less business. The key factor is whether the brand has already been on the customer’s radar during the decision-making phase.

An example: Someone is considering switching from a car to a bike more often in everyday life. In the past, this person would likely have conducted several Google searches, read blog posts, compared reviews, and worked their way step by step toward a purchase decision.

Today, research using an AI system might look like this: The person first asks, in general terms, whether a bike can be a real alternative to a car for everyday use. Then they ask which type of bike is suitable for roads, forest trails, and gravel tracks. At this point, a gravel bike might come up for the first time. In the next step, the potential customer asks about good brands for gravel bikes. This is where it becomes relevant for manufacturers: Is their own brand mentioned or not?

The more specific the decision becomes, the more classic search channels like Google or YouTube come back into play: reviews, first-hand user experiences, price comparisons, and purchases typically still happen via Google.

This creates a major opportunity: with GEO, I can be present early in the customer journey — exactly where users make their decision.

GEO gives small brands the opportunity to gain attention in a market dominated by big brands.

The reason for this is the so-called query fan-out. In the past, a search query was sent directly to the index. For the keyword or prompt “what are the best running shoes,” there would then be a search results page dominated by strong domains, large online shops, and well-known brands. Today, AI intervenes in three steps:

• Context analysis: AI tries to understand the implicit question, not just the words in the typed text.
• Fan-Out: The original question is broken down into many synthetic search queries. These also include topics and details that weren’t explicitly mentioned in the query.
• Aggregation: The results of these many synthetic queries are combined and summarized by AI into what is known as an “AI Overview.”

So, behind the simple search query “what are the best running shoes” many detailed questions are created in the background. Not only does Google try to understand what I’ve typed, it also tries to understand what I mean, what I need and what answer I might expect. This generates synthetic search queries based on what Google knows about me. The result is no longer a more or less neutral set of search results, but a highly personalized answer.

This means that even small brands or websites with lower visibility can gain exposure through expertise in specific niches. It’s no longer about ranking for a single main keyword—where large brands are virtually unbeatable anyway—but rather about being mentioned in relation to one or more of the many synthetic search queries.

In other words: In the past, SEO for Google was a competition for the Grand Trophy—the number-one ranking. Today, it’s more like a raffle. Every specific, detailed question that’s thoroughly answered on your website is a ticket in the pot. The fan-out is the draw. In a way, AI draws a ticket from the pot for every synthetic search query. The more tickets you’ve put into the pot, the more chances you have of gaining visibility. That’s why content further down the funnel can be especially valuable: content that offers specific expert knowledge or covers specific subtopics in depth.

Incidentally, the same applies to YouTube: AskYouTube—the new AI feature currently being rolled out to Premium subscribers—answers users’ questions directly and links to the relevant video with a timestamp. In doing so, YouTube explicitly prioritizes relevance over reach.¹ This means that with outstanding content, even small channels have a chance.

The third opportunity is particularly exciting for B2B marketing: Traditional search queries don’t reveal who the user is or what exactly they need a solution for. For example, users search Google for “tax advisor Vienna” or, more specifically, “tax advisor for establishing a GmbH.” However, these keywords provide only limited insight into who is searching. That’s why SEO strategies often focus on service pages: What services do we offer? What products are available? What subtopics can we cover, or how can we group products to address as many search terms as possible? All competitors then optimize for these obvious terms, and in the end, many providers end up competing for the same spots in the top 5 search results.

This is a particular problem for B2B. The keywords don’t make it clear whether the search is about personal use or whether someone is looking for a solution for a large company. That’s why SEO for B2B has always been a real challenge, with a lot of wasted reach.

In AI search, people use prompts to define who they are, describe their situation, explain the problem they want to solve, and specify which conditions matter. Instead of just typing “tax advisor Vienna” into the search bar, users write a prompt like: I’m a freelance web designer with an annual revenue of 60,000 euros. Recommend 5 affordable tax advisors near me who specialize in digital service providers. Or: I’m OMV and I’m looking for a new tax advisory firm.

This is precisely where the great potential lies: In addition to traditional “What” pages about services and products, you can also create “For Whom” pages that address specific target groups and use cases.

Business models designed solely to generate as much traffic as possible and monetize that traffic are in a tough spot today. GEO is not the new source of traffic either. If your business depends only on clicks, there is not much to be optimistic about. Many AI answers provide the information that users would have previously searched for on websites—without them having to make a single click: How tall is St. Stephen’s Cathedral? When was the first moon landing? For questions like these, the AI draws on its training data, leaving hardly any room for optimization.

GEO primarily comes into play when AI searches the web: When does the next WordCamp Vienna take place? I’m stuck at the airport in Doha—how do I get out of here? What’s the best vinyl version of Ben Webster Meets Oscar Peterson? To answer questions like these, AI needs up-to-date or specialized information and performs a web search.

And this is exactly where it gets interesting for website operators. If a website covers a topic comprehensively, provides up-to-date information, and offers specific expert knowledge, the chances increase that AI systems will use it as a source.

The real shift, however, is happening on an even deeper level: When AI systems decide which brands to mention in their responses, it’s no longer only visibility in search results and rankings that matter, but also mentions, brand awareness, and sentiment. For example, a gravel bike manufacturer should be mentioned when a user asks about the best brands for gravel bikes—and ideally in a positive way.

This is exactly where GEO comes in: A website should not only be optimized for classic Google rankings, but also qualify as a source for AI responses and be mentioned by AI systems as a solution or a trustworthy provider.

To learn how this can be implemented in practice, read our article on the 3 most important GEO measures.