Over the past few months, Tencent's stock price has been puzzling. From a high of HK$639 in early 2026, it fell to around HK$472 by the end of April, a drop of over 25%. During this period, there were no obvious negative developments for Tencent: gaming revenue remained stable, the advertising business was recovering, and WeChat's monthly active users continued to exceed one billion. Each earnings report was more impressive than the last, yet the stock price kept declining. Analysts offered various explanations, but they often felt superficial—correct yet missing the core issue.

A new perspective emerged after seeing a screenshot from a friend's phone. It showed iPhone battery usage for the day, with data for two apps side by side: WeChat, with 2 hours and 1 minute of screen-on time, consuming 50% of the battery; and Claude, with 2 hours and 11 minutes of screen-on time, consuming 26%. The usage durations were nearly identical. My friend added a note: "I'm now spending more time on Claude than on WeChat."

While this single screenshot doesn't prove a macro trend, it offers an observation point: in one specific user's day, an AI application is competing directly with WeChat for attention. Not TikTok, not Weibo, not any product that has challenged WeChat over the past 15 years, but an entirely different type of application.

Capital markets are pricing in this structural unease. The issue for Tencent is not about today's earnings, but about user time flowing toward a domain where WeChat has never truly established a defense. To understand what this means, we must start from a more fundamental proposition: in an era of information overload, what is the truly scarce resource?

Attention as a Scarce Resource In 1971, Herbert Simon wrote in a frequently cited yet often misinterpreted paper that an abundance of information necessarily leads to a poverty of attention. His logic was concise and powerful: information consumes the attention of its recipients. Therefore, the more information there is, the scarcer attention becomes. This insight, proposed before the internet's birth, found its full historical weight in the mobile internet era. In 1997, Michael Goldhaber built on this to propose the concept of the "attention economy": in the information age, attention itself is currency. Whoever captures user attention can monetize it—whether through advertising, subscriptions, or data. The business models of internet platforms are, in essence, the monetization of attention.

This framework provides the first key to understanding WeChat's competitive history. Over the past 15 years, every challenge WeChat faced was superficially a battle of product features but fundamentally a reallocation of attention share: telecom operators lost the time and money users spent on SMS; Weibo lost the idle time users spent in public squares; TikTok competed for users' fragmented sensory time during waits, commutes, and before sleep.

But attention is not homogeneous. Herbert Simon, a cognitive scientist, understood that human cognitive resources exist in hierarchies—superficial perceptual attention and deep, thoughtful attention are qualitatively different. This distinction becomes crucial in the AI era, a point we will return to later.

From SMS Killer to National Operating System To understand WeChat's competitive history, one must first grasp how disruptive its emergence was. When WeChat launched in 2011, China's mobile communication market was in the golden age of telecom operators. SMS was the main channel for daily communication, MMS was the precursor to social media, and Fetion was China Mobile's failed attempt to build a closed ecosystem. WeChat, with its minimalist instant messaging tool,颠覆了 this order. Leveraging QQ's user base for rapid growth, it used features like "Shake," "People Nearby," and "Moments" to transform a communication tool into a social platform. This降维打击 against telecom operators took less than three years.

WeChat's first real competitors were other instant messaging products. Momo, launched in 2011, focused on stranger social networking and一度 showed fierce user growth. Miliao, a同期 product from Xiaomi with Lei Jun personally endorsing it, was seen as WeChat's most dangerous early rival. However, neither Momo nor Miliao could break through the network effects barrier WeChat had already established.

Network effects are an extremely残酷的 competitive mechanism. Economists Carl Shapiro and Hal Varian systematically explained this logic in their 1999 book "Information Rules": when a product's value increases non-linearly with the number of users, the user base of the first mover itself forms an entry barrier.

More critically, communication products face "bilateral lock-in": you need to convince not only users to switch platforms but also all their contacts simultaneously. This coordination cost is, in practice, almost an impossible task. Miliao最终 ceased service in 2019; Momo found its own niche but never threatened WeChat's core position.

The second wave of challenges came from content platforms. The rise of Weibo once led outsiders to believe the main battlefield of social media was in public squares, not private messaging spaces. The launch of WeChat Moments in 2012 was, to a significant extent, a strategic response to Weibo—using a "semi-closed" circle of friends to counter Weibo's public broadcast logic. This proved to be an extremely correct judgment. The intimacy and information quality of Moments were better suited to the social habits of mainstream Chinese users than Weibo's information deluge. Weibo subsequently gradually transformed into a public domain platform for celebrities and media, coexisting with WeChat in a differentiated manner rather than engaging in direct confrontation.

The third wave came from short video. TikTok launched in 2016, with Kuaishou slightly earlier. This was the first genuine time分流危机 WeChat encountered. Short video doesn't compete for communication or social relationship chains; it directly competes for every minute users spend idly in front of their phone screens. TikTok's algorithmic recommendation mechanism is extremely precise; it is essentially the ultimate engineering of attention economy logic: using reinforcement learning algorithms to continuously optimize the "attention capture rate per minute," keeping users scrolling for an hour without expectation.

Platform economist Jean Tirole's theory of two-sided markets has an interesting application here: WeChat and TikTok are actually competing for two sides of the same users—as content consumers and as communication nodes. TikTok only challenges the former; it does not and cannot challenge the latter. This explains why WeChat, despite some losses during the short video era, did not see its foundation shaken.

WeChat's response was Channels. Launched in 2020, its development was slow initially, but under Zhang Xiaolong's持续 strategic bet, Channels' daily active users and usage time have increased year by year. By 2024, Channels had become a non-negligible third force in China's short video领域. WeChat's reaction this time cannot be called a failure, but it was certainly not an overwhelming victory. TikTok's advantage in user time remains largely unshaken.

Before AI, Did WeChat Ever Lose? Frankly, before AI, WeChat never lost any competition in a fundamental sense. It achieved complete victory over telecom operators. It achieved complete victory over early instant messaging rivals. It achieved strategic non-defeat against Weibo (the two diverged, but WeChat's core social functions remained unshakable). Against short video, it experienced局部失守; user time spent on TikTok was indeed eroded, but WeChat's reliance is not on entertainment time but on its irreplaceable position as communication infrastructure.

This is WeChat's true moat: it is not an app you like to use; it is an app you have to use. Work groups are on WeChat, payment QR codes are via WeChat Pay, parents contact you through WeChat, school notifications come via WeChat. Ordering food delivery, making medical appointments, checking housing provident fund—all can be done through WeChat Mini Programs. WeChat long ago ceased to be just a communication tool; it is the underlying operating system of digital life in China.

Here, a theoretical concept worth辨析 is switching costs. Economists distinguish between two types: functional switching costs (the cost of relearning and adapting to a new product) and relational switching costs (the cost of disconnecting from an existing social network when switching platforms). WeChat's moat is primarily the latter. Users are on WeChat not necessarily because it is the best, but because social networks are already established there in practice. The cost of leaving is not losing a tool but losing a set of social connections.

David Evans and Richard Schmalensee, in studying platform competition, noted that once this kind of relational lock-in forms, there are almost only two ways to break it: either the entire social network migrates collectively (extremely difficult) or a new demand scenario emerges that doesn't require migrating old relationships (possible).

AI恰好 provides the second possibility.

AI: The First Challenger That Truly Alters Time Structure Returning to Herbert Simon's hierarchy of attention. In the past, all of WeChat's competitors were vying for users'浅层注意力—entertainment time, fragmented browsing time, social chat time. This time could be taken by TikTok or occupied by games, but WeChat's "essential communication time" remained unbreakable.

The emergence of AI assistants has, for the first time,触动了一种不同性质的时间: deep attention time. When users open DeepSeek, Claude, or GPT, what are they doing? Writing articles, researching information, analyzing data, processing legal documents, learning new knowledge, conducting investment research. The common feature of these scenarios is that users are in a state of active cognitive engagement, not passive sensory consumption. This is what Herbert Simon termed "thoughtful attention,"其稀缺程度和价值密度 far exceed the fragmented time spent scrolling through Moments.

From the battery screenshot, one user spent 2 hours and 11 minutes interacting with Claude in a single day, exceeding their active time on WeChat. This is not entertainment substitution; it is workflow substitution. Workflow, however, is territory WeChat has never truly competed for.

Tencent is certainly aware of this threat. Its Hunyuan large model has been launched, and WeChat internally is exploring integration paths for AI features. But there is a fundamental tension here: WeChat's product philosophy is "use and leave." Zhang Xiaolong has emphasized on multiple occasions that a good product should not make users沉迷; it should allow them to complete tasks and leave quickly. This conflicts intrinsically with the product logic of AI assistants, which aim to become users'深度工作伴侣.

Beyond the conflict in product philosophy, there is a more direct reality: Hunyuan's actual performance has been underwhelming. During the most intense period of the global AI race from 2024 to 2025, OpenAI, Anthropic, Google, Meta, and even China's本土 DeepSeek were iterating at a肉眼可见的速度, with capability curves rising steeply. Hunyuan remained lukewarm, rarely appearing in the top tier in domestic mainstream evaluations and having almost no international presence.

For a company with massive user data, ample funds, and top engineers, this is a result that puzzles the market. Tencent is not losing on resources; it is more likely losing on priorities and organizational culture: whether a company with social and gaming in its基因 can muster sufficient strategic will in an era requiring "all in" on foundational model R&D remains an open question.

A deeper issue is that competition among AI assistants is not about features but about竞争 for cognitive trust. Economics has the concept of "experience goods"—products whose quality cannot be judged before consumption, only known after use. AI assistants go a step further: they are highly personalized experience goods. The longer you use them, the deeper the context, the more accurately they understand you, and the higher the switching cost becomes. Once this stickiness forms, its mechanism is identical to the social relationship chain stickiness WeChat built years ago, but it binds not users' social relationships but their cognitive habits and working memory.

Tencent's situation is somewhat similar to China Mobile's in 2011—not that the product is bad or resources are insufficient, but that a new competitive dimension has been defined, and you are not the one defining it.

The Final Outcome of Time Returning to the initial question: where is the time going? Herbert Simon's proposition has been fully validated in the mobile internet era: attention is indeed the scarcest resource. Over the past 15 years, WeChat won battle after battle in the attention war. It first took communication time from telecom operators, then used Moments to secure social time, and further used Mini Programs and Pay to lock in service time. WeChat built a super platform that几乎全天候复盖 the digital lives of Chinese users.

But the hierarchical structure of attention means this war never truly ends. The rise of AI assistants represents a new paradigm of time consumption—not娱乐式的浅层注意力消费, but工具式的深层注意力投入. Every minute users spend on AI is actively completing a task: writing, researching, analyzing, learning. The value density of this time is far higher than scrolling through Moments or watching Channels. This high-value time恰恰 is territory WeChat has never truly guarded before.

Shapiro and Varian made a prophetic judgment in "Information Rules": in the information economy, lock-in effects are the core source of lasting competitive advantage, but the basis for lock-in can shift with changes in technological paradigms. WeChat locks in social relationships; TikTok locks in sensory habits; AI will lock in cognitive habits and working memory. These three types of lock-in target different levels of the human attention system and do not have simple substitution relationships with each other.

Tencent is not incapable of responding: it lacks neither funds, data, nor user base. But the challenge it faces is at the product philosophy level: WeChat's greatness lies in knowing what it is and what it is not. The rise of AI assistants is重新画定 "what is the core of digital life." If the answer shifts from "social" to "thinking and work," WeChat's product positioning will require a difficult self-reconstruction, for which there is no precedent.

The stock price likely reflects precisely this uncertainty. The market is not worried about how much Tencent earns today but about where WeChat will stand in the next hierarchical migration of user attention.

This question cannot be answered by a single battery screenshot. But it is the right question to ask.