Summary of Key Points
In recent months, major internet companies (such as ByteDance, Tencent, and Alibaba) have made a strategic shift: they have moved away from investing billions in C-side AI products (like DouBao App, Alibaba QianWen App, and Tencent YuanBao) at the beginning of the year to focus on the AI office software market. Although the initial investment in C-side products brought short-term traffic spikes, they faced issues with poor user retention, low conversion rates, and high costs associated with using large AI models (which consumed significant computing power), significantly impacting profits. The office software market, on the other hand, presents a clearer opportunity for revenue generation due to the clear payment intentions of businesses and the frequent, essential needs of companies for tasks such as document creation, data analysis, and code generation.
1. The Change in Cost Logic: Why Didn’t the C-side Investment Model Work?
In the mobile internet era, the strategy of “investing to gain traffic and then using traffic to generate advertising revenue” was effective because the more app users there were, the closer the marginal cost approached zero (e.g., by spreading server costs over a larger user base). However, AI models are entirely different. Every time a user asks a question or generates content, it actually uses GPU power, and the more users there are and the more frequently the models are used, the higher the costs become. Moreover, there is no network effect (unlike with social platforms like WeChat, where more users make the service more valuable).
Financial reports clearly illustrate this issue: Tencent’s new AI products dragged down its profits by 8.8 billion yuan in the first quarter; Alibaba’s AI lab division lost 4.2 yuan for every 1 yuan it earned. Billions of investments only resulted in short-term traffic spikes (for example, DouBao’s daily active users once reached 145 million), but once marketing efforts stopped, users disappeared, and retention and conversion rates fell far below expectations. The C-side model simply wasn’t sustainable, forcing these companies to change their approach.
2. The Collective Shift to AI Office Software: The Key Reasons for This Choice
There are three main reasons why companies are turning to AI office software:
1. High-frequency and Clear Needs: Businesses need to create documents, spreadsheets, PPTs, analyze data, and write code daily—these are repetitive and time-consuming tasks.
2. Willingness to Pay: Companies are willing to pay if AI can replace manual tasks and improve efficiency (for example, if an AI can generate a report that would normally take an hour in just 5 minutes). This makes the revenue more stable compared to the free habits of C-side users.
3. Clear Delivery Standards: It’s easy for businesses to assess the quality of generated documents and the functionality of code, unlike C-side AI chat services, where the focus is more on the experience rather than the actual outcome.
In short, the office software market represents the closest opportunity for AI to generate real revenue.
3. Major Organizational Reorganizations: How Are Companies Consolidating Their AI Efforts?
Previously, internal AI teams within these companies were competing with each other (for example, ByteDance had multiple products like DouBao, TRAE, and Kousi; Alibaba had three office AI products). This led to resource dispersion, unnecessary duplication of computing power, and internal competition for the same customers. Now, companies are consolidating their efforts:
- ByteDance: The Feishu team, AI programming team TRAE, and the intelligent platform Kousi have all been integrated into the DouBao system.
- Alibaba: The three office AI products have been combined into “QianWen Office,” which is now under the responsibility of the DingTalk CEO.
- Tencent: The QClaw product center has been incorporated into the WorkBuddy system.
The goal is to centralize AI capabilities into a single product that can be used across different software platforms and even continue to function after the device is turned off, effectively turning the AI into a “digital employee” that integrates into business processes.
4. Challenges to Revenue Generation: What Hurdles Remain in AI Office Software?
Despite the right direction, there are still significant challenges to overcome before achieving profitability:
1. Unestablished Payment Habits: While the SaaS market in the U.S. is mature, Chinese companies are still hesitant to pay for AI tools, as many are accustomed to using free alternatives.
2. Integration into Business Processes: Companies buy AI to reduce costs and increase efficiency, not just for novelty. For example, AI systems must be integrated into specific business processes (such as financial reimbursement and customer management) to be valuable. This requires manufacturers to understand the industry, provide customized solutions, and ensure data security.
3. Quality Stability: Generated documents and code must be error-free and reliable; otherwise, companies won’t use the AI services.
From “pilot projects” to actual profitability, companies need to prove that AI office software can truly help them save money and improve efficiency.
In Conclusion
The shift from C-side investment to AI office software by these major companies is a rational response to market trends. However, to truly generate revenue, they must overcome three key hurdles: establishing payment habits, deeply integrating AI into business processes, and ensuring the quality of their products. Only those companies that can successfully make AI a necessary tool for businesses will emerge as winners in the long run.