The Challenge
Until just a few months ago, the website was the digital center – info hub, business card, shop, application. The most important entry points to the website were usually through:
Search engines (Google, Bing)
Direct URL entry
Advertising, social media, newsletters
The first two entry points – search engines and direct URL entry – are increasingly under pressure. Today, these two entry points account for between 30 and 60% of traffic, depending on the industry and website. And this is exactly where AI systems are increasingly intervening. Instead of redirecting users, they deliver answers directly – faster, more precisely, more contextually. Without clicks, without website visits, at best with a reference to our source.
In the past, users would have had to click through multiple websites – today, a single question to an LLM (Large Language Model) suffices. The content that was once read on our site now appears directly in the LLM's response window. This changes everything.
A Concrete Example
A user asks ChatGPT: "What funding is available for renovating an old farmhouse in the Canton of Bern?"
ChatGPT responds directly: "Various foundations such as Swiss Mountain Aid or the Canton of Bern's Heritage Conservation come into question. It's important that the property is located in a mountain area and is relevant for heritage conservation."
Then follows a follow-up question: "Does Swiss Mountain Aid support my project in Thun?"
Answer: "Since Thun is not located in the recognized mountain area, funding through Swiss Mountain Aid is rather unlikely."
Conclusion: Relevant information is delivered without clicks – precisely, helpfully, and without ever opening a website. Studies show measurable impacts: Ahrefs documented a 34.5% decline in click-through rates for position-1 rankings when AI Overviews are present, while for searches with AI Overviews, the zero-click rate increases from about 60% to an average of 83%. But what can we do?
Human Inspiration – Machine Readiness
The future of websites must focus on two layers:
1. Experience Layer – for Humans
Goal: Trust, identity, emotion – in seconds. The focus is on impact, not information depth. The homepage becomes a stage and brand magnet. The rest remains functional.
Strong hero sections with video, 3D, creative headlines, fascination
Maximum of 3 clearly staged user journeys
Stories, cases, or products with emotional, fascinating impact
2. Knowledge Layer – for Machines
Goal: Structure content so that various LLMs can use it. Focus on function and information depth, not on impact or emotions. This area can have an application-like, sober character.
Highly optimized content area for blogs, studies, events, and use JSON-LD, as LLMs rely heavily on this
Semantic HTML, schema.org, structured data
Implement AI-based search and optimize content so that LLMs can crawl, understand, and cite it (e.g., question-answer areas like FAQs)
Implement question - answer - source structure
No overloaded JS frameworks without server-side rendering (e.g., SPAs without SSR)
The Advantages of the New Focus
Clarity instead of overload: Experience layer emotional, knowledge layer searchable
AI-ready: Content is found – and passed on
Less maintenance effort: Clear separation, efficient maintenance
Technically compatible: Clean foundation for APIs & integrations
Interesting aspect: While zero-click searches reduce traffic, initial data from Ahrefs shows that visitors from AI-powered search convert 23 times better than those from traditional search. Less traffic, but higher quality.
You can see how this concept is applied on the newly launched website of the Radio-Oncology Center

Webseite of the Radio Onkologiezentrum Biel