For a long time, the logic was clear: people search for information, Google shows links, users visit websites, read content, form an opinion and decide. A significant portion of what we today call digital marketing is based on this chain - from search engine optimization (SEO) to content strategy to lead generation.
With the rise of dialogue-based AI systems, this very chain is changing. Not dramatically, largely quietly - but noticeably and measurably. When someone has a question today, they don't necessarily go through the classic Google search. Increasingly, the question goes directly to ChatGPT, Claude, or Gemini. But traditional search results are also changing: answers are being integrated directly into search results pages. This means that fewer and fewer human visitors navigate to the website that originally provided the information (zero-click search). These days, only about 20 percent of all search queries lead to a direct referral to a website - and this trend is declining further. Mobile searches are particularly affected here. This is understandable, as accessing relevant information on mobile devices requires significantly more time and coordination effort due to device-related accessibility constraints.
For companies, this is more than a technical advancement. It's an architectural shift that redefines the role of websites, brands, and content. I've described how this development extends far beyond marketing and reveals something more fundamental about how humans process knowledge in a separate essay. Here, the focus is on what this shift means in practical terms.
Google doesn't lose its significance
Since the rise of modern language models, the question regularly arises whether traditional search engines are losing relevance. The answer is more complex than many headlines suggest. No, Google hasn't disappeared from the market - contrary to many predictions. The search engine continues to process enormous volumes of queries, with a distinctly upward trend. At about 190,000 search queries per second, that's almost three times as much as in 2020.
As the following animations show, the percentage decline in clicks on search results doesn't mean fewer referrals in total, since overall usage has increased significantly in recent years.
However, the way people access information is changing. Increasingly, knowledge is no longer created through:
- multiple search queries,
- visits to many web pages,
- or comparing different sources,
but through a single dialogue-based process with an AI system.
Google itself is also integrating this approach into its traditional information model. The search engine was originally a navigation system for the web. It showed links - people left Google and visited websites. Today, Google is increasingly becoming an answer machine, integrating answers from its own AI directly into search results through so-called 'AI Overviews'.
It has always been Google's goal to bring users as 'close' as possible to the information they seek. It's hard to get closer than integrating answers directly into search results ... unless we're talking about direct implantation of knowledge (and thinking) into the human brain in the distant or not-so-distant future.
What this means for websites
Even though the total number of visitors Google refers has increased, current developments already show far-reaching consequences for some industries, and at least from a future perspective for others, for websites, content marketing, and digital visibility. Because if the website merely acts as a content supplier for externally delivered information, then the website as such, as a 'packaging' for the brand, serves no function. Or does it?
On closer inspection, we find that its importance and function are shifting. And it would probably be a fatal mistake to assume you could now simply forgo your own website.
Yes, language models change decision-making processes. Whether in pre-selection or advising on products with specific characteristics. Because sometimes AI can find a suitable selection faster. Or provide unbiased advice, while people try to push their products and don't act in the buyer's interest (for example, with financial products). When people are overwhelmed, when they mistrust other people, then using an LLM makes sense. These systems provide easier access to transparency against this background.
So what task remains for the traditional website? Even if part of the decision-making process is outsourced to AI systems, the website remains a crucial space for trust. With its brand interface, it represents a robust validation instance. It is a citable source with a concrete, unsmoothed value proposition.
Against this background, the importance of traditional content strategies must be questioned: do glossaries, FAQ pages, knowledge sections, and practical tips still have a reason to exist? After all, comparable and individually tailored information can be obtained through the LLM.
Visibility remains decisive - but it's changing
This doesn't mean Google or domain visibility becomes irrelevant. Quite the opposite. However, in this context, clicks, rankings, reach, and visitor traffic are no longer the only performance metrics. Another layer is added: websites are increasingly becoming part of the knowledge space from which AI systems draw, synthesize, and evaluate information.
This means: in the future, a website won't just compete to be found by humans. It will also compete to be captured and considered by AI-based systems. Optimization for 'machines' recalls the early days of traditional search engine optimization - except the machines today are different and apply different criteria.
Optimization is thus shifting increasingly away from pure focus on key terms and keywords toward:
- semantic clarity,
- structure,
- citability,
- authority,
- brand trust,
- and demonstrable credibility.
High-quality texts are therefore more important than ever if you want to be cited or recommended by AI-powered search systems. This reality rejects the quick and superficial generation of content. Content created exclusively with LLMs not only lacks any differentiation potential but can permanently damage a brand.
It is therefore important to enrich content with aspects that AI cannot synthesize: human elements like experience, originality (not conformity), personal style, clear positioning - and also friction and resistance. This applies to specialist articles as much as to brands, experts, and entire companies. To remain visible in a world where machines can synthesize the majority of average content, you must deliver above-average, differentiable, and thus outstanding quality.
Conclusion
Three consequences can be drawn from all this for brands and companies:
- First: the question 'Will we be found?' is no longer enough. It must be expanded to 'Will we be cited?' and 'Will we be recommended?' - by people and by machines.
- Second: generic content is losing value. What used to bring reach - explanatory texts, general guides, list articles - is increasingly being delivered better, faster, and more individually by AI systems.
- Third: brands become more important, not less. In a world where information is synthesized, smoothed, and decontextualized, what gains importance is what cannot be easily synthesized - a recognizable voice, clear positioning, a distinct profile. That's exactly what creates trust: with readers and with AI systems processing this content.
The quality of content not only still has its place. In the sea of consensus uniformity, it's more important than ever when it comes to convincing people and machines. In marketing, corporate communications, in life. It's about thought leadership. And that can only be credibly communicated if you as a company take content responsibility - not an outsourced 'thinking service'. Websites and digital visibility remain the first and most important contact points with new customers. Regardless of whether these are mentioned by an LLM or appear in Google search results. For most business models, it doesn't matter whether your website is integrated at the first or second position in the decision-making process. The main thing is that it's included at all.
Make sure your website becomes or remains the most important reference and contact point. Make it a thought leader; not bland and characterless content miscellany. Or worse, an obvious AI parody ...