AI Search instead of Google clicks: How businesses can stay visible now

If AI provides the answers – why do people still visit your website?

Google AI Overviews, ChatGPT, Perplexity, Gemini and other AI systems are transforming the way people search for and find information. Increasingly, users are receiving their answers directly within the search engine or via an AI chatbot, without even visiting a website. For businesses, this represents a fundamental shift: website traffic alone is becoming less meaningful. Instead, the focus is shifting to the quality of each individual visit.

But what does this mean in practical terms for businesses, organisations and public bodies? What demands will this place on content, search engine optimisation and digital user experiences – and how can website managers respond to them?

These questions were the focus of our LinkedIn webinar with Søren Schaffstein and Svea Landschoof. Below, we’ve summarised the key takeaways for you. Enjoy!

AI Search is transforming the customer journey

For many years, searching for information on the internet followed a similar pattern: users would enter a few keywords into Google, click on a search result and work their way through various websites until they had found the information they were looking for. Today, this process is looking increasingly different. Instead of individual keywords, users are formulating complete questions or even complex tasks, such as:

‘Plan a peaceful summer holiday in Sweden, with plenty of nature, lakes, small villages and as little tourism as possible. The destination should be easily accessible by car from northern Germany.’

AI systems such as ChatGPT or Gemini provide a personalised travel suggestion within a matter of seconds – often including an itinerary, places of interest and recommendations. The user receives their result without having to visit numerous websites or piece together information themselves. This change in search behaviour leads to what is known as a ‘zero-click journey’: the search ends before the user even clicks on a website.

This therefore raises a new question for website managers: no longer „How do I attract as many visitors as possible?" but „How do I help the visitors who actually come to the site to find what they’re looking for as quickly and reliably as possible?"

Fewer visitors – but significantly more valuable visitors

A fall in click-through rates does not automatically mean a decline in success. On the contrary: anyone who visits a website today, despite the use of AI search, is often already pursuing a specific goal. These users may, for example, wish to:

• make a decision
• compare offers
• get in touch
• book a service
• obtain detailed information

Each individual visitor therefore has significantly greater potential than was the case just a few years ago. At the same time, expectations are changing. People who are used to receiving a clear answer from AI systems within a matter of seconds expect the same efficiency on company websites. Lengthy navigation paths, confusing page structures or content that is difficult to understand are perceived as obstacles much more quickly.

SEO remains important

There are currently repeated claims that traditional SEO has had its day. In fact, the opposite is true. A technically sound website, high-quality content and a good information architecture remain the foundation for visibility on the web. At the same time, however, new requirements are emerging: whilst traditional search engines primarily evaluate and rank links, AI systems attempt to understand content, categorise it and summarise it into a direct answer. That is why two terms are becoming increasingly important:

• Answer Engine Optimisation (AEO): Optimisation for AI-powered answer systems.
• Generative Engine Optimisation (GEO): Optimisation of content for generative AI systems.

The aim is no longer simply to appear as high as possible in the search results. Rather, content should be structured in such a way that AI systems can understand it, categorise it and incorporate it into their answers as a trustworthy source. Do you need help implementing this? Get in touch with our SEO experts.

Get advice from our SEO experts

What AI systems expect from good content

User queries rather than corporate jargon

Many websites are written from a corporate perspective. They describe products, services or organisations – often using internal terminology. Users, however, search differently: they formulate specific questions.

Instead of using only headings such as ‘Our Services’ or ‘Product Portfolio’, content should be geared more towards typical user queries. FAQ or Q&A sections are a good example of this: questions and answers are clearly linked, making it easier for both users and AI systems to understand the content.

The most important answer first

Many websites begin with long introductions before getting to the actual information. For users and AI systems, the opposite is often more effective: the key answer should appear as early as possible, followed only then by background information, examples or further details. This so-called ‘answer-first’ principle improves both readability and the likelihood that content will be interpreted correctly by AI systems.

A clear structure rather than walls of text

• Logically organised headings
• A clear page hierarchy
• Well-organised internal links
• Semantically structured content
• Thematically related content clusters

These measures do not just help search engines and AI systems. Above all, they make it easier for visitors to find their way around.

Thematic authority is becoming increasingly important

Rather than covering as many different topics as possible in a superficial manner, AI systems are increasingly favouring websites that demonstrate expertise and inspire trust within a clearly defined subject area. This includes, amongst other things, up-to-date content, verifiable sources, practical examples, customer testimonials, user reviews and regular updates. In short: quality beats quantity!

UX and MX are converging

In the past, search engine optimisation and user experience were often viewed as separate concepts: SEO ensured visibility, whilst UX ensured that visitors could find their way around the website. With AI Search, these two areas are increasingly merging. Added to this is another, still emerging aspect: what is known as ‘machine experience’ (MX) – in other words, the question of how well machines can understand and process content.

In practical terms, MX means, for example, that content is tagged with structured data (such as Schema.org markup) so that AI systems can unambiguously identify prices, opening hours, authors or product attributes. A solid technical foundation is also essential: fast loading times, an HTML framework accessible to crawlers rather than content rendered purely on the client side, and consistent, machine-readable sitemaps. The clearer these signals are, the more reliably AI systems can interpret and cite content correctly.

A modern website must therefore meet three requirements simultaneously: it must be easy for people to understand, it must be technically sound, and it must present information in such a way that AI systems can interpret it unambiguously. The good news is that many measures address all three areas at once.

When users arrive at the website, every second counts

Particularly large websites face an additional challenge today. Local authorities, universities, associations and large companies often have thousands of subpages. Although the information people are looking for does exist, it is often difficult to find.

Whilst users used to be willing to click through several levels of navigation, this willingness is steadily declining. The key question is therefore no longer how to increase traffic, but how to guide visitors to the right information as quickly as possible.

Practical example from the webinar: During the webinar, Svea Landschoof presented the rek.ai platform, which is specifically designed to address the challenge described above. The focus is on content-rich websites – such as those run by local authorities, universities, associations, tourism organisations or companies with extensive information resources.

Personalised user experiences

Rather than displaying the same content to every visitor, rek.ai analyses various contextual details such as the device being used, the time of day, the visitor’s location, their previous browsing behaviour and their interests. Based on this, content is dynamically prioritised and relevant recommendations are displayed.

A practical example comes from the Swedish municipality of Uddevalla. There, the quick links on the homepage automatically adapt to the typical behaviour of comparable users. If, for example, a particularly large number of parents access the website via their smartphones on a Monday morning, the link for reporting a schoolchild’s illness can automatically be given a more prominent position. Seasonal trends can also be taken into account. When, during a cold spell in Sweden, a particularly large number of citizens were searching for information on frozen water pipes, the system recognised the increased demand and automatically gave greater prominence to the relevant content.

Smart recommendations instead of rigid navigation

As well as personalised quick links, context-based content recommendations can also be displayed. At the end of an article, visitors are automatically shown further content on related topics – much like on streaming platforms or online shops.

For example, if someone on the Visit Sweden website shows an interest in walking holidays, the recommendations on the homepage will subsequently change accordingly and display further content relating to walking routes or nature experiences. Universities and local authorities can also use this principle to guide visitors to relevant further information and reduce bounce rates.

Semantic search rather than simple keyword search

Traditional search functions often only deliver good results when users enter the exact right term. Modern semantic search systems, on the other hand, recognise the user’s actual search intent: they understand contextual relationships, take similar terms into account and deliver relevant results even when different phrasing is used or typos occur. This is particularly useful for multilingual websites. Users can formulate search queries in their own language without having to translate them first.

Automatically generated questions and answers

Another key feature of rek.ai is its automatically generated FAQs. Suitable FAQ entries are suggested based on existing page content, which editorial teams can then review, edit and approve.

This approach deliberately follows the ‘human-in-the-loop’ principle: content is not published without being checked. Instead, AI supports the editorial team in the creation process, whilst the responsibility for subject-matter verification remains with humans. This results in structured content that makes it easier for both visitors and AI systems to navigate the site.

Easy integration into TYPO3

For TYPO3 projects, dkd provides its own extension that allows the functions of rek.ai to be integrated with ease. Once set up, features such as personalised recommendations, intelligent search suggestions and automatically generated FAQ sections can be incorporated without the need for time-consuming bespoke development. As the solution is script-based, it can in principle also be used in other content management systems. For TYPO3, however, the extension significantly simplifies setup and configuration and enables seamless integration into existing projects.

The website remains central – its role is changing

AI Search will not replace the traditional website. However, its role is undergoing a fundamental shift. Anyone wishing to be successful in the future should not focus solely on increasing visitor numbers. What matters far more is the quality of the user experience, how quickly visitors can find what they are looking for, and whether content is presented in a way that is understandable to both humans and AI systems.

Three first steps for your website

• Content and SEO audit: Check your key pages against the ‘Answer-First’ principle and ensure they focus on users’ questions rather than corporate jargon.
• Technical and GDPR check: Check whether structured data and a clean, crawlable site architecture are in place.
• Navigation test and tracking: Ask test users to search for a specific piece of information and measure how many clicks it takes them to find it.

Making your website AI search-friendly