Summary of Key Points
Feishu has been split and integrated into Doubao by ByteDance, with its lead, Xie Xin, being demoted two levels. On the surface, this seems like the result of AI replacing traditional software, but there are three underlying reasons:
1. Feishu required significant investment with low revenue, failing to meet ByteDance’s strategic expectations;
2. Doubao’s enterprise version needs Feishu’s native integration capabilities to build a competitive edge;
3. Most fundamentally, ByteDance has realized the reality of China’s ToB (Business-to-Enterprise) market—working with large companies is both time-consuming and unprofitable, while scaling up with smaller businesses aligns better with ByteDance’s internet strategy.
Detailed Analysis
1. Demotion of Feishu: Not Due to AI Competition, but Because It Didn’t Meet Expectations
As a strategic division of ByteDance, Feishu invested approximately 4000 people over 7 years, incurring annual labor costs of 3.2-4 billion yuan, yet its annual subscription revenue was only 3 billion yuan, resulting in a severe imbalance between input and output. ByteDance had been holding onto Feishu because it hadn’t found a more suitable entry point for the ToB market. However, the rise of Doubao (ByteDance’s AI product) presented a new opportunity, leading to Feishu’s strategic repositioning. In simple terms, Feishu wasn’t generating enough revenue to sustain itself, and with Doubao showing more promise, it has shifted from a “leading role” to a “supporting role.”
2. Why Does Doubao Need Feishu? Native Integration Is the Key Competitiveness Factor
The core challenge for enterprise AI products (such as Doubao’s enterprise version) is not the technology of large models, but the ability to deeply integrate with enterprise tools. For example, Workbuddy can only check inventory when integrated with Kingdee’s ERP, whereas Kingdee’s own AI can directly generate order fulfillment strategies (placing orders, making purchases) because they are “natively integrated” rather than relying on simple API calls. Without Feishu’s integration, Doubao would only be a “diminished version” of Workbuddy. With Feishu, Doubao can quickly develop AI products that effectively handle enterprise collaboration and processes.
3. ByteDance’s Insight into the ToB Market: Large Companies Are a Challenge, Small and Medium-Sized Businesses Are the Profitable Targets
Feishu previously focused on the strategy of “advanced organizations using Feishu first,” similar to SAP’s approach of targeting large companies before expanding. However, this didn’t work in China for several reasons:
- Complex Business Models: Large companies have diverse needs and high requirements for delivery teams, but the high cost of such teams makes projects prone to delays and losses (e.g., a 3-million yuan project might end in a loss);
- Powerful Clients: Clients often change requirements or adjust project architectures mid-way, imposing additional costs on vendors;
- Limited Payability: Similar projects might generate 20 million yuan for SAP/Oracle, but only 3 million yuan for Chinese companies, with intense competition.
In contrast, small and medium-sized businesses have more standardized needs. For instance, Kingdee’s subscription business has a gross margin of 96.5%, generating profits from day one. ByteDance has realized that working with large companies is a slow and costly process, not suited to its “scalable and rapid expansion” approach.
4. Is the Splitting of Feishu a Good Move? Using ByteDance’s Advantages to Break into the ToB Market
Some argue that splitting Feishu was a mistake, but it depends on the perspective:
- From a traditional ToB perspective, Feishu’s team’s expertise in serving large companies makes the split a loss;
- From ByteDance’s perspective, the split is beneficial. ByteDance excels at creating scalable products (like Doubao), which has a large consumer base that can easily convert into enterprise orders. Using its internet approach, it can quickly build a profitable ToB business. This isn’t a abandonment of Feishu; rather, it’s about leveraging Feishu’s resources (users, experience) to support Doubao’s AI strategy and pursue a more suitable ToB path.
Conclusion
Feishu’s demotion is not the end of its journey but a shift in ByteDance’s ToB strategy—moving from focusing on large companies to using AI and scalability to serve small and medium-sized businesses. This reflects ByteDance’s deep understanding of the Chinese ToB market: rather than struggling in the complexities of large companies, it’s better to use its strengths to create a ToB business that can be quickly replicated and profitable.