Three months after launch, Doubao Work has reached 8 million users and set a revenue target of 2 billion yuan. We see AI office products shifting from answering questions to handling tasks, with vendors increasingly using user scale and commercial targets to test genuine demand.

What this is

Doubao Work is an AI product designed for office scenarios. It does more than answer questions: it attempts to understand user intent, invoke tools, and complete multistep tasks such as organizing materials, generating documents, and processing data. These agents—AI systems that can independently plan steps and call software to achieve goals—are becoming enterprises’ new entry point for AI.

Eight million users represent substantial adoption, but “users” cannot be equated directly with paying customers or enterprise seats. The 2 billion yuan figure is closer to a commercialization target than realized revenue; the two should not be conflated.

Industry view

We believe Doubao Work sends a positive signal: AI office products have moved beyond competing solely on model capabilities and are entering a phase centered on workflows, collaboration, and delivery. Platforms with models, search, and enterprise services are better positioned to embed AI into high-frequency work.

The risks are equally clear. Office tasks often involve company files, customer data, and permission systems. Any information leak, inaccurate output, or operational error will lead buyers to impose significantly stricter security requirements. Whether 8 million users develop stable paying habits also remains to be seen; if most usage comes from free trials, the 2 billion yuan target may take longer to achieve.

Impact on regular people

For enterprise IT: Companies will increasingly evaluate whether AI can integrate with existing software, permissions, and audit systems, rather than judging only how fluent its answers are. Procurement will gradually focus on deployment costs, data security, and actual hours saved.

For individual professionals: More document organization, information retrieval, and repetitive processes will be handed to AI. People’s core work will shift toward checking results, making judgments, and taking responsibility for business outcomes.

For the consumer market: AI applications charging for memberships, task packages, or cloud storage will proliferate, while users will pay closer attention to whether results justify the price. Wider free adoption does not necessarily mean every feature will remain free indefinitely.