Summary of Key Points
This article discusses how China's leading internet companies have proactively "reinvented themselves" with the advent of the AI era. Even when their existing businesses are still profitable and at their peak, they have chosen to launch new AI products (such as intelligent assistants) to replace or transform their old ones in order to adapt to the new market cycle. Using cases from companies like DingTalk, Lark, and Tencent, the article reveals the key to their survival: the willingness to "kill" their past selves at the height of their success, allowing new businesses to emerge from within.
Why Do Large Companies Need to "Kill" Their Most Successful Businesses?
You might wonder why they would dismantle profitable operations. The answer is simple: technology and user needs evolve too rapidly, and waiting until the old businesses fail would be too late.
The article provides three classic examples:
- Tencent WeChat replacing QQ: In 2010, QQ was a nationally used app, but Zhang Xiaolong believed it belonged to the computer era and needed a mobile counterpart. Instead of having the QQ team work on it, Tencent organized internal competitions among three teams, which led to the birth of WeChat. Today, WeChat is the mainstream, and although QQ still exists, Tencent has survived the mobile internet era thanks to WeChat.
- Alibaba's Mobile Taobao replacing the web version: In 2013, Taobao's web version was dominant in e-commerce, but Jack Ma decided that the mobile team should replace it. Two years later, Mobile Taobao became the main platform for major shopping events like Double Eleven, and the web version gradually faded into obscurity.
- ByteDance's TikTok replacing Jintoutiao: In 2018, Jintoutiao was still a popular app, but TikTok surpassed it. ByteDance changed its company name from "Jintoutiao" to "ByteDance," making TikTok the new core of the business.
These examples show that no matter how successful an old business is, it cannot stop the tide of new technology. Only by actively replacing it while it is still profitable can a company survive.
The Challenge of DingTalk's "Wukong" Transformation
DingTalk's story is the most dramatic. Wu Zhao, the founder, started with a failed social app called "LaiLai" and found success by addressing executives' concerns about employee communication and task progress. He added features like read/unread notifications, attendance tracking, and approval processes, which made DingTalk a leader in enterprise collaboration tools.
However, with the arrival of AI, DingTalk's traditional model (manual operation of functions) became obsolete. In 2025, Wu Zhao was brought back to transform DingTalk into an AI-powered tool called "Wukong," allowing AI to directly utilize DingTalk's underlying capabilities without manual intervention. His management approach was radical:
- Strict attendance requirements (9 a.m. check-ins, morning and evening meetings, mandatory work at 1:15 p.m. after lunch),
- Management staff were required to learn Python, and the technical team had to analyze code volume,
- Product managers were expected to visit three companies per week, using only DingTalk for external communications.
This strict management led to internal dissatisfaction, and Wu Zhao resigned in June 2026. However, DingTalk's AI transformation continued with Wukong being integrated into Alibaba's Qianwen office suite.
The challenge is not just changing the product but also the organization and culture, which may require sacrificing certain aspects of the company's existing structure.
Why Lark, Despite Success, Needed to Be Reformed?
Lark had revenue exceeding 3 billion yuan in 2025 and grew by 100% in the second quarter of 2026, with 90% of new customers purchasing its AI products. Despite its growth, ByteDance decided to restructure it:
- The product team was merged into ByteDance's AI assistant "DouBao," and Lark's CEO reported to DouBao's lead,
- The sales and marketing teams were integrated into ByteDance's cloud services platform, Volcano Engine.
The reason? ByteDance's strategy had shifted: AI assistants are seen as the future core. As an office tool, Lark needed to be integrated with DouBao to become more intelligent. For example, combining Lark's spreadsheets with DouBao's data analysis capabilities would create greater value.
This shows that even if a business is thriving, it must be adjusted if it does not align with the company's future strategy. Office tools that do not integrate with AI will gradually be phased out as AI becomes mainstream.
The Race for AI Agents
Leading companies are focusing on developing AI agents (intelligent assistants), such as Tencent's WorkBuddy, Baidu's "Dazhi," and Meituan's CatPaw. These agents are more than just chatbots; they can directly access company resources to perform tasks:
- Tencent WorkBuddy: Ma Huateng actively participates in its development, providing ample resources, potentially making it the third strategic product on par with QQ and WeChat.
- Baidu's Dazhi: The internal assistant team was merged, and Lark's content was integrated into its systems.
- Meituan CatPaw: Its models were integrated into the delivery system to manage orders for hundreds of thousands of riders and millions of merchants.
What they are competing for is the "future entry point": AI agents can replace traditional software, becoming a bridge between users and corporate resources. The company that develops the best agent will gain a competitive advantage in the AI era.
The Cost of Self-Revolution
The article draws a parallel with Lao She's "Teahouse," where Wang Lifa tried to improve his teahouse throughout his life but failed because he did not change its fundamental operations. Corporate self-revolution is more intense and comes at a higher cost:
- Wu Zhao used strict management to push the AI transformation of DingTalk, eventually leading to his resignation.
- Lark's CEO, Xie Xin, had her reporting structure changed from directly to Liang Rubo to DouBao's lead, indicating a shift in power.
- Within Tencent, multiple AI teams competed, with some being merged.
The lesson is that self-revolution is not just about slogans; it requires genuine effort to change products, organizations, and even personal interests. Only by doing so can companies survive in the face of change.
In Conclusion
Over the past two decades, the most brutal lesson in China's internet industry has been not about defeating competitors but about asking: "When your most successful business begins to hold you back, do you have the courage to eliminate it first?" To survive, you must be willing to reinvent yourself.