Summary of Key Points
This news article focuses on Meituan's AI transformation: Wang Puzhong, the CEO in charge of Meituan's core local business, analyzes the "four mismatches" that hinder the successful implementation of AI in companies today. He shares four stages of change within Meituan from a strategy of "spending money to trial and error" to actually generating value. The article also reveals Meituan's specific practices in the field of AI, such as its applications in the pharmaceutical industry and the release of AI tools, as well as the management's attitude towards AI transformation—viewing it as a "must-do" task without reckless investment.
Detailed Breakdown and Interpretation
1. Why is the implementation of AI in companies so underwhelming? What are the "four mismatches" mentioned by Wang Puzhong?
Although AI technology is everywhere, most companies fail to see tangible benefits despite their investments. Wang Puzhong believes the problem lies in four areas where things do not align:
- Cognitive mismatch: Either AI is overhyped as a panacea (capable of solving everything) or it is completely dismissed as useless, leading to extreme views.
- Efficiency mismatch: High-end large models are used for trivial tasks (like writing ordinary weekly reports) rather than on the most critical business aspects (such as increasing sales or improving customer service).
- Scenario mismatch: AI is still in a peripheral role within companies (e.g., using chatbots for customer service) and has not been integrated into core operations (such as supply chain optimization or targeted marketing).
- Evaluation mismatch: Managers are eager for immediate results without considering whether the company's internal processes and employee capabilities can keep up with the pace of AI development.
These mismatches result in wasted funds that do not translate into actual productivity.
2. Meituan's own AI journey: Four steps from trial and error to generating value
Meituan's AI transformation has not been smooth, but it has gone through four stages:
- Stage 1 (February-March): Everyone was involved, leading to high costs
The company encouraged everyone to work on AI projects (referred to internally as the "shrimp farming campaign"), resulting in substantial expenses. However, AI-generated errors disrupted normal operations (e.g., providing incorrect strategies to merchants).
- Stage 2 (April onwards): Establishment of an AI organization for structured efforts
Specialized AI teams were formed within various departments, shifting the focus from chaotic experimentation to organized work.
- Stage 3 (June-July): A competitive approach to find effective solutions
By using a competitive framework, Meituan realized that AI transformation requires collaboration among business, organizational, and technical teams. For example, drug-selling teams needed to understand how to use AI, the company's structure had to support its implementation, and tech teams had to develop suitable tools.
- Stage 4 (July onwards): Integration into processes and value creation
AI was finally integrated into internal product workflows, starting to generate actual benefits.
3. Meituan's "secret" to successful AI transformation: Focus on high-value scenarios and a three-pronged approach
Wang Puzhong emphasizes that companies should not approach AI blindly. Two key points are essential:
- Start with high-value scenarios: For instance, Meituan began piloting AI in the pharmaceutical industry by sending experts to help chain pharmacies like Shuyu Pingmin optimize their operations. They also made the AI platform CatPaw available for pharmaceutical businesses to test, as this sector requires precise services.
- Integration of business, organization, and technology: AI cannot be developed solely by the tech department; business teams must understand how AI can solve their problems, organizational structures need to adapt (e.g., establishing AI teams), and tech teams must create tools that fit business needs.
4. Meituan management's attitude towards AI: A "must-do" task with targeted investment
Meituan's leadership has a clear stance on AI:
- Wang Xing (Founder): AI transformation is a mandatory requirement, not an optional choice. All companies should use AI in their products (e.g., AI features in apps) and work processes (e.g., using AI for report writing), but investments must be within the company's financial capabilities to avoid reckless spending.
- Chen Shaohui (CFO): This is the era of AI, and Meituan wants to participate in cutting-edge technological advancements while making rational investments.
This year, Meituan has accelerated the release of AI products: the AI browser Tabbit was tested publicly in March, the AI community "Miyou" was launched in May, the "Run errands Skill" was made available, and the large model LongCat-2.0 was released in June, along with the AI tool "Xiaotuan Health Butler." Their actions are both proactive and cautious.
5. Practical applications of Meituan's AI: Already generating revenue for merchants
In addition to internal use, Meituan's AI is also serving external businesses:
- The first batch of AI experts helped pharmaceutical companies like Shuyu Pingmin optimize inventory and recommend products.
- The comprehensive AI platform CatPaw has been piloted with these businesses to manage their operations.
- In the first half of the year, "Xiaotuan Health Butler" was launched to provide users with AI-based health consultation services.
These applications all start from high-value scenarios, directly addressing real problems for both merchants and users, rather than being mere showpieces.
Conclusion
Meituan's story of AI transformation shows that companies should not follow the trend without a clear plan. They need to address cognitive, efficiency, scenario, and evaluation mismatches and focus on high-value areas. Only by integrating business, organization, and technology can they turn AI into a real source of productivity. Meituan's journey from trial and error to generating value serves as a valuable example for other companies to follow.