Author: Wildcard

In the first quarter of 2026, the monthly active users of the Baidu App dropped to 655 million, a decrease of 10% year-on-year.

Baidu is not an ordinary app; for the past 20 years in the Chinese internet, "Baidu it" has almost become synonymous with "searching online." It is the starting point of traffic, the entrance to information, the "tap" on which countless websites rely for survival.

However, in the AI era, users are increasingly inclined to obtain answers directly from AI assistants rather than by clicking links on search engines.

The world shares the same cool and heat.

On the other side of the Pacific, the same script is playing out in a more intense manner. A "traffic collapse" affecting the entire internet is taking place.

And people are indifferent.

01 No One Survives

In March 2026, the analytics company Chartbeat exclusively provided a set of data to Axios that ignited the entire publishing industry.

This dataset tracked the traffic changes of thousands of client websites under Chartbeat over the past two years, leading to a shocking conclusion — small publishers (websites with an average of 1,000 to 10,000 page views per day) saw their search referral traffic plummet by 60%. Medium-sized publishers fell by 47%. Large publishers experienced a relatively "mild" decline as well, but still had a drop of 22%.

The smaller the scale, the harsher the death. The reason is not complicated — large media have brand recognition, apps, email subscription lists, and loyal users who visit directly. Small publishers have almost all their fortunes tied to Google search, and when search traffic falls off a cliff, they have no buffer whatsoever.

Page views from Google Search decreased by 34% between December 2024 and December 2025. Google Discover also saw a 15% decrease during the same period. The entire Google "traffic tap" is being tightened.

But for individual websites, the numbers are far more brutal than these averages.

A report released by Loopex Digital in May 2026 listed the websites most affected. Business Insider's traffic plummeted by 48.4%. WebMD's monthly visitors dropped from 122 million to 69.5 million. The world's largest programming Q&A community, Stack Overflow, saw a decline of 35.5%, with Google search volume decreasing by 32.4%. Quora dropped by 28.1%. Even image resource sites like Unsplash were not spared, with monthly users falling from 26.8 million to 17.7 million, a drop of 34%.

If you look at this list horizontally, you will find a disturbing pattern — the hardest hit are precisely the core components of the "internet information infrastructure" over the past decade.

News media (Business Insider, CNN), knowledge Q&A platforms (Quora, Stack Overflow), medical information sites (WebMD), image resource platforms (Unsplash) — these websites share a common characteristic: their fundamental value lies in answering users' questions.

And now, there are others answering for them.

02 The "Zero-Click" Era

To understand what is happening, one needs to clarify a concept — "Zero-click Search."

This term refers to: when a user inputs a question on a search engine, they get an answer directly on the search results page without clicking any links or visiting any websites. Approximately 60% of searches worldwide now end in zero clicks. On mobile, this number is as high as 77%.

Zero-click search did not just appear in the AI era. Google has long been doing the same thing with features like "Featured Snippets," "Knowledge Panels," and "People Also Ask" — displaying answers directly on the search results page, reducing users' reasons to click away from Google. But AI Overviews have pushed this to an entirely different magnitude.

According to various research data, Google AI Overviews currently cover approximately 48% of search queries, a year-on-year increase of 58%. When an AI Overview appears on a search results page, the click-through rate for the top-ranking organic search result decreases by 34% to 58% — depending on which study you refer to, but the trend is consistent.

Seer Interactive's analysis is even more aggressive, suggesting that in queries with AI Overview, the overall click-through rate for organic search has plummeted by 61%.

In simple terms: your website is still in the search results, and your ranking may not have changed at all, but your click volume may have been halved. This is the most confusing part for many website operators — "I didn't do anything wrong, yet the traffic suddenly dropped off a cliff."

But Google AI Overviews is only half the story.

The other half is a whole emerging "AI search" ecosystem that is thriving wildly.

ChatGPT now processes over 1 billion search queries weekly. Perplexity's monthly query volume also exceeds 1 billion. Claude, Gemini, and Copilot are all rapidly growing — Goodie's AI search traffic report for 2026 shows that Claude's AI referral share surged from 11.8% to 18.5% in early 2026, the fastest growth among all AI search engines.

The operational logic of these AI products fundamentally differs from that of traditional search engines. Google's past business logic was "to send users to websites" — you search for a question, Google gives you a string of links, you click in, and Google's job is done. This is a "traffic directing" model.

But the logic of AI chat products is "to give you the answer directly." It reads your content, understands your knowledge, synthesizes an answer, and then delivers it directly to the user. The user is satisfied, the user leaves, and throughout, they do not need to know where the answer came from, nor do they need to visit any website.

This is why merely focusing on changes in Google traffic is insufficient. What is truly happening is far deeper than "Google algorithm updates."

The entire internet is transitioning from a "link network" to an "answer network."

For the past 20 years, the basic flow of information on the internet has been: creators produce content → search engines index and rank → users click links → visit websites → content creators gain traffic and ad revenue. This is a closed loop. Each participant has its own role and benefits.

Now, this chain is being severed. The new chain has become: creators’ content is captured and learned by AI → AI directly answers users' questions → users leave satisfied → content creators gain nothing.

The middle "click" process, which has supported the entire digital publishing industry, digital advertising industry, and content marketing industry for 20 years, is being systematically eliminated.

03 China's "Parallel Universe"

Some may think this is largely a story about Google and the English internet, having little to do with China.

On the contrary. China is experiencing the same thing, only the companies taking the traffic are different.

In the U.S., the main force driving traffic away is Google's own AI Overviews — Google places an AI-generated summary at the top of its own search results, keeping users on the search results page and no longer leading them to external websites. This is equivalent to Google "cutting" its own traffic.

The situation in China is more intense. The traffic is not being taken by traditional search engines themselves but by a whole batch of independent AI-native applications like Doubao, Qianwen, and DeepSeek, which are directly snatching users from traditional search.

The numbers speak volumes.

QuestMobile's data shows that in April 2026, the average monthly usage frequency of search engine apps per user dropped 18.8% year-on-year, and usage time decreased 11.8%. Meanwhile, during the same period, the monthly active users of AI-native apps approached 500 million, with the average monthly usage frequency reaching 91 times and average usage duration of 180 minutes. By June, Doubao's active users had already surpassed 528 million.

528 million — this number is approaching Baidu App's 655 million monthly active users. Furthermore, the former is still increasing while the latter is decreasing.

QuestMobile's May 2026 survey discovered that among its observed 25 industry apps, 28% faced a year-on-year decline in both user usage frequency and duration, while another 40% saw at least one metric decline. Search, automotive information, and travel were the three industries that first felt the impact of AI information distribution. Among online travel industry users, 69.4% were also using AI-native apps. For automotive information users, this ratio was 51.1%.

There is a particularly thought-provoking piece of data. After the content of Pacific Auto was cited by Doubao, Qianwen, and DeepSeek, it reached a total of 12.37 million users. This sounds like good news — until you realize this number is equivalent to 113.8 times the self-generated traffic of Pacific Auto App.

113.8 times. In other words, over a hundred times more users "consumed" the content of Pacific Auto through AI platforms, but these users never set foot on the Pacific Auto website. They asked a question about buying a car in Doubao's chat window, and Doubao quoted Pacific Auto's content to provide an answer, leaving the user satisfied and closing the conversation.

Pacific Auto contributed knowledge, the AI platform garnered users, while Pacific Auto gained nothing.

The companies taking traffic are different, but the logic being eroded is exactly the same — whether it is Google AI Overviews or Doubao, they are fundamentally doing the same thing: reading the knowledge of content creators, synthesizing answers, and directly delivering them to users, bypassing the original source.

04 One-Way Traffic

At this point, some may think: It's true that AI has taken traffic from search engines, but don't AI products also reference sources and provide links? Aren't they also "driving traffic" to websites?

Theoretically, yes. In practice, it is almost negligible.

Chartbeat's data shows that from December 2024 to December 2025, traffic from AI chat tools did indeed increase by over 200%. But even after this increase, it accounted for less than 1% of the total page views for publishers.

Contentsquare provided even more precise numbers in its 2026 Digital Experience Benchmark Report — AI referral traffic skyrocketed 632% year-on-year but still only accounted for 0.2% of all traffic. Moreover, the quality of that 0.2% isn't optimistic, with 53.6% of AI-referred users leaving directly after.

600% growth, 0.2% share, with over half leaving directly.

This set of numbers paints a very harsh picture — AI is consuming a large amount of traffic from search engines, but it is not "forwarding" this traffic to content creators. It is simply keeping the traffic on its own platform.

What’s even more interesting is the behavior of users after they click to a website from AI.

Multiple analyses indicate that many users who came via AI chat tools actually just want to verify whether the AI's answer is accurate — they open the webpage, glance at it, confirm that the AI did not make nonsense, and then close the page and leave. They are not there to "consume content," but rather to "verify facts." This type of visit holds little commercial value for content creators.

This creates a huge "ecological inequality."

AI companies use the content of publishers and content creators to train their models. When users ask questions to AI, the AI utilizes this acquired knowledge to generate answers. Users leave satisfied on the AI platform. Advertising revenue (if any) remains on the AI platform. Meanwhile, those who created the original knowledge and content receive almost nothing.

This is not merely a problem of traffic allocation. It is shaking the very foundation of the entire digital content economy.

For the past 20 years, the basic logic of digital advertising on the internet has been built on the chain of "visit → display → click." Users visit websites, see advertisements, and advertisers pay. When the very starting point of "visit" begins to disappear, the foundation of this entire chain ceases to exist. For media and content platforms that rely on advertising revenue to survive, this poses an existential threat.

Gartner predicts that by 2028, organic traffic from traditional search engines will decline by 50%. Not by 5%, not by 10% — but by half. If this prediction holds true, we will live in a completely different internet from today.

05 The Deepest Paradox

This may be the most profound paradox of this "traffic cut."

Returning to the beginning with Baidu. Baidu is using AI to compensate for the decline in search advertising — in the first quarter of 2026, AI-driven business revenue reached 13.6 billion yuan, a year-on-year increase of 49%, accounting for more than half of core business revenue for the first time. While Baidu is using AI to kill the old Baidu, it is trying to recreate a new Baidu with AI.

But the bigger problem does not reside with Baidu. At least Baidu has the capacity to transform. What about the countless small and medium websites, content platforms, and vertical communities that have been sustained by traffic from Baidu? They have no capability to implement AI, nor do they have the brand to attract direct visits from users. When search engines no longer direct traffic and AI does not either, what will they do?

For 20 years, there has been an unwritten "contract" between search engines and websites: you produce content, and I give you traffic. This contract has no formal agreement, no legal guarantee, but is based on a simple commercial logic — search engines need good content to attract users, and websites need traffic from search engines to survive. Both sides benefit.

Now, AI has disrupted this equation. Search engines (and AI platforms) realize: I don’t need to send users to websites; I can provide the answers myself. Users even prefer this way — no clicks required, no waiting for page loads, no sifting through a pile of ads to find the actual content.

A product that offers greater customer satisfaction is precisely one that causes more pain for content creators.

When in an ecosystem, the enhancement of user experience and the loss of creator benefits become two sides of the same coin, that ecosystem reaches a crossroads where it must be redesigned.

AI is killing the ecosystem that feeds it.

And what the internet will look like next remains to be seen; no one has provided a suitable answer yet.