Summary of Key Points
Recently, both Alibaba and ByteDance have undergone significant internal integrations of their AI-based office products: Alibaba merged three overlapping intelligent agents into a single product called “Qianwen Office” and launched a paid public beta, while ByteDance incorporated the Lark product team into the DouBao team, which is now managed by the person in charge of DouBao. Although the approaches differ (Alibaba is moving from collaboration tools to AI solutions, whereas ByteDance is integrating AI assistants with office products), the goal is the same—to eliminate internal product conflicts, unify user access points, and advance commercialization. Both companies recognize a common consensus: in the era of AI-powered offices, users will only accept one primary platform, and whether enterprises are willing to pay for it depends on whether the AI truly enhances work efficiency.
1. Alibaba: Three Products Competed, Now Unified with Charging
Alibaba’s new Qianwen Office is the result of combining three internally similar intelligent agents (QoderWork, Wukong, and MuleRun). These products had overlapping functions and were competing for resources within the company, so integration was necessary. The unified Qianwen Office now offers services to both businesses and individuals, with a monthly fee of 98 yuan for the individual version and 198 yuan per user for the enterprise version; the free version allows only 10 concurrent tasks.
Interestingly, Alibaba has not charged for the Qianwen App, despite previously promising not to. The reason is that the Qianwen App is designed as a chat assistant for general users, and offering it for free helps attract them. In contrast, Qianwen Office is a productivity tool for work scenarios, and businesses and professionals are more willing to pay for efficiency-boosting features. Although they share the “Qianwen” brand, there is a risk of user confusion (e.g., thinking they are different versions of the same product), which Alibaba aims to address.
2. ByteDance: Lark Team Integrated into DouBao to Resolve Decision-Making Issues
ByteDance faced similar issues to Alibaba: the AI functions in Lark (Aily) overlapped with those in DouBao’s enterprise version, leading to resource conflicts. For example, although DouBao’s enterprise version incorporated Lark’s document functionality, it could not read the document content during internal testing due to data discrepancies.
At the end of July, ByteDance merged the Lark team into DouBao, with Zhao Qi taking over as the new manager, and the former Lark leader Xie Xin reporting to him. This move aims to make DouBao the primary platform for AI-powered offices, with Lark serving as a provider of basic functions such as documents and meetings. However, while organizational integration is easy, aligning product data is more challenging. It was discovered that DouBao still cannot read Lark’s documents, a problem that needs to be resolved gradually. Fortunately, DouBao has a large user base (382 million monthly active users), which gives it an advantage as the primary platform.
3. Why Both Companies Agree on “Only One Entrance”?
Both Alibaba and ByteDance understand that the key to AI-powered offices is not the number of features but being the first choice for users. For example, if employees have to choose between DingTalk and Qianwen Office or Lark and DouBao when writing reports, it reduces efficiency.
The ideal future office workflow would be to open an intelligent agent (such as Qianwen Office or DouBao) and simply request something like, “Organize last week’s meeting minutes and send them to the clients in a spreadsheet.” The agent would automatically use meeting data, document tools, and email systems to complete the task without requiring users to open multiple apps. Therefore, there must be a single, unified entrance; otherwise, users will find it cumbersome, and the products will not be successful.
4. The Key to Commercialization: Are Enterprises Willing to Pay?
The companies are now charging because their internal integrations are complete. Charging before resolving conflicts would mean supporting products that have not yet proven their value, leading to user confusion. With a unified entrance, pricing becomes more straightforward.
However, whether they can actually make money depends on whether enterprises see the value in their AI tools. The logic for enterprise customers is simple: does the improved efficiency justify the monthly cost? For instance, if an employee uses Qianwen Office and completes 2 hours more of work per day, the monthly fee of 198 yuan would be justified. If the tool merely seems useful but doesn’t provide real benefits, user renewal rates will be low. This is similar to hiring employees; the investment must be deemed worthwhile for it to be included in the budget.
5. New Changes in Future Office Software
In the future, collaboration tools may become secondary components of AI-powered offices. For example, Lark documents might no longer be accessed directly but instead used as part of a larger system called DouBao; DingTalk’s communication features could be integrated into Qianwen Office for information synchronization.
Collaboration tools will shift from being the primary interface to becoming backend services. This is an inevitable trend in AI-powered offices, as users want solutions to problems, not just tools to operate. The company that successfully implements this transition will gain a foothold in the AI office market.