For several decades, digital media and search engines operated according to a simple chain: information was published on a website; when users searched for a few keywords, the search engine offered a list of links; readers clicked on the link they preferred, visited the website, and perhaps read other stories there as well.
Each link in this chain had its own value. Visibility turned into visits, visits into an audience, and the audience into revenue, subscriptions or other forms of engagement. The entire system was built around one simple model: people had to come to your website for you to be able to turn their attention into value.
Today, that mechanism operates according to a different logic. Search engines are increasingly trying not simply to find and display the source of information, but to answer the user’s question directly. AI-powered systems can gather facts from different sources, synthesize them and present a single, coherent answer without requiring the user to visit the original source website.
As a result, a new situation is emerging: journalism can reach readers without bringing them to a news outlet’s website.
This is the most important consequence of “zero-click” information consumption. And its impact goes far beyond declining visits: it is changing the model of the relationship between media and audiences on which a significant part of the digital media economy has been built over the past decades.
And this raises the most important question: what happens when the reader and information meet through this new model?
Finding the reader is no longer enough – you also have to “bring them back”
The numbers make this shift particularly tangible. A PPC Land article reports that, in July 2026, the 100 largest U.S. media publishers spent an estimated $113 million on paid search, according to Similarweb tracking data reported by Adweek. That spending was up 41% year over year and 274% compared with July 2023. The number of visits generated by paid search reached 23.7 million, up 39% from the previous year.
The numbers point to a rather striking picture. For years, publishers have relied on search engines to bring readers to their websites. Now, some major media publishers are spending increasingly large sums to buy back that search traffic. And this rapid growth in spending is taking place at a time when the search environment is undergoing a fundamental transformation, with organic search referrals declining and the old path to reaching readers becoming increasingly uncertain.
What if people consume your journalism but you never see them?
This is where the idea of zero-click search becomes more interesting than simply a technical term. A person may read the results of your journalism without you ever seeing that reader. They may know the facts you gathered without visiting your website. They may even use an answer built on the work of your journalists without knowing that a particular media outlet is behind that work.
This situation is particularly significant for the media because a news organization may lose more than a single visit. It may lose the reader’s path back to itself.
When someone visits a website, a media outlet has an opportunity to build a relationship with that person. They may read one story, then another, subscribe, sign up for a newsletter, download an app or simply return the next day.
If another system provides the person with all the information, the first and most important part of that relationship takes place somewhere else. That is why the problem is much bigger than website traffic statistics.
The media is already changing its behavior
A survey conducted by Digiday and Arc XP, involving 85 publisher respondents, shows how publishers are trying to adapt to this shift.
Seventy-six percent of respondents said they are seeing a shift toward summary-first consumption. Sixty-one percent reported a decline in direct visits to publisher websites, while 47% said AI search has already affected referral traffic to their sites.
As a result, publishers are looking for new ways to reach and retain audiences beyond search.
The same survey shows that media organizations are already changing their approach at several levels. Eighty-five percent of respondents said they are actively optimizing content for AI discovery or citation, while 76% said they have changed their SEO or content structure because of AI search. At the same time, 62% said they are working with AI search companies, while 76% said they are licensing or syndicating their content to AI platforms.
At the same time, interest in owned channels is growing. Sixty-seven percent of respondents have increased their social media presence, while 52% are investing more in owned channels such as newsletters, apps and podcasts.
Publishers seem to understand that it is no longer enough simply to be present where people search. They also need a connection with readers that does not depend on someone else’s search engine.
The question is not only technological
All of this could be viewed simply as a matter of the relationship between media organizations and technology companies. But there is another change, one that is more human.
People do not only search through AI systems. They talk to them, ask them questions, sometimes thank them, and sometimes return to the same conversation. In other words, something that was initially created as a tool can gradually become part of a person’s everyday behavior.
And this is where one recent study adds an interesting new layer to the story.
When a chatbot is more than just a tool
A study published in August in the academic journal Technology in Society surveyed 7,027 people from Germany, China, South Africa and the United States.
The authors, Genia Kostka and Hui Zhou, sought, among other things, to understand what kind of relationships people develop with general-purpose chatbots, including systems such as ChatGPT and DeepSeek. The study examines the prevalence of emotional attachment to chatbots, as well as the factors shaping it and its consequences.
The findings are striking. Among chatbot users, 48.06% consider chatbots their friends. Some 61.69% report being polite to them, for example by saying thank you. And 35.05% say they miss their chatbots after several days without using them.
Another 39.57% say they feel less lonely. Some 57.27% say they feel free from the discomfort of being judged by others when interacting with chatbots, while 67.55% report a sense of privacy.
Perhaps the most striking figure is this: 20.94% say they would rather share their deepest secrets with a chatbot than with friends, family members or other trusted people.
Of course, these figures cannot be taken to mean that people are preparing to replace their human relationships with AI.
The study itself does not draw such a simple conclusion. On the contrary, the authors found that people with larger social networks may also show stronger emotional attachment to chatbots. In their interpretation, the relationship people develop with AI can, in some cases, complement human relationships rather than simply replace them.
This finding matters for another reason. It shows that the power of AI lies not only in its speed. It is available, it does not get tired, and it does not interrupt. It can continue the same conversation. And for some people, it creates an environment in which they feel more at ease.
These same characteristics are becoming important from the perspective of information consumption as well. When someone asks, “What happened today?”, “What does this statement mean?”, “What is happening in this country?”, or “What is the difference between these two viewpoints?”, AI stops being simply a new form of search. It becomes an intermediary between people and information.
And here, the most important question for the media may no longer be how many people saw its headline, but whether the media outlet was present where people went to seek information.
What should a news organization do?
The answer cannot simply be, “We just need to do better SEO.” At the same time, the metrics used to measure audience growth need to be reconsidered. Alongside website traffic, the number of returning readers, direct visits, subscriptions, newsletter and app usage, and how often a media outlet appears as a source in AI-generated answers are becoming increasingly important. In other words, media organizations need to build several paths through which readers can discover, use and return to the same news outlet.
At the core of the content, however, the same requirement remains: it must be verified, accurate and clearly sourced. The spread of AI does not change that standard, but it does change the environment in which the results of that work need to be accessible and recognizable.