ByteDance restructures Feishu into the Doubao ecosystem

Text | Beyond the Layout, Author | Hua Hua

This may be the most important day in Feishu's ten-year history.

On the morning of July 30, ByteDance sent an internal email announcing a major reorganization of its structure:

This means that starting today, Feishu will no longer operate independently and will no longer have separate sales and marketing teams.

A decade-long independent business ended with a single email.

Many people would think this at first glance. After all, Feishu is one of ByteDance’s most heavily invested innovative businesses. In 2023, Liang Ruibo himself stated that the R&D investment in Feishu and VolcEngine is no less than that in Douyin and TikTok.

A business that invested no less than TikTok has been completely dismantled and rebuilt today. What else could this be called but a loss?

But a glance at the timeline reveals a different story.

For several years, Feishu has been waging an independent war. It aims to prove itself as China’s Office, China’s Slack, China’s Notion, China’s Microsoft 365...

It's too eager to become a super gateway.

What’s the result?

Google has integrated Gemini into Workspace. Microsoft has embedded Copilot throughout the entire Office experience. Tencent has directly embedded AI into WeCom and Tencent Docs. Alibaba has extended Qwen to cover office scenarios.

Today, leading global companies almost simultaneously took the same action: abandoning entry-point logic and embedding AI directly into workflows.

This also made Feishu suddenly realize that the platform logic it had upheld for years was becoming increasingly difficult to sustain.

What happened today was not the acquisition of Feishu, but rather Feishu's first reduction from an independent product to an infrastructure capability.

It's not Feishu that lost; it's AI that won. AI won against everyone, including what Feishu itself aspired to become.

Looking back at Feishu's 2024 organizational streamlining, which shifted from casting a wide net to focusing on top-tier initiatives, many interpreted it as a contraction. But it wasn't.

Perhaps at that point, Feishu realized that selling tools alone wouldn't win this battle—it needed to embed itself within a larger, more fundamental organizational framework.

Today, the framework has arrived. But it wasn't found by Feishu itself—it was chosen by ByteDance.

The answer is hidden in a set of data.

As of June 2026, the Doubao large model surpassed 18 trillion daily token requests. VolcEngine is now a leading player in China’s public cloud MaaS market. Based on the average consumption in July, ByteDance’s large model business generates an annualized revenue of $4 billion, exceeding the combined annualized revenue of all other model companies in China.

What about Feishu?

In Q2 2026, over 90% of new Feishu customers simultaneously purchased Feishu AI products.

This number is what matters. It shows that enterprise customers have already voted with their feet—they’re not here to buy Feishu, they’re here to buy AI.

Feishu is merely a channel through which AI reaches enterprises.

Bytes must act at this node. The reason is simple: AI has become the sole entry point for enterprise procurement.

Enterprises do not purchase office software and cloud platforms separately, nor do they merely buy large model APIs; they purchase an integrated AI solution.

Whoever can bundle AI capabilities, office tools, and cloud infrastructure to serve customers well will secure enterprise orders.

DouBao, Feishu, and VolcEngine—only when combined can they compete.

ByteDance acknowledged for the first time in this email that AI is not a business line, but an infrastructure layer.

You cannot operate infrastructure as a conventional department—you must reorganize the entire organization around it.

This aligns perfectly with the highest-level strategy of the byte.

In June, at the Volcano Engine Power Conference, ByteDance CEO Liang Ruibo stated that scaling the heights of AI is currently ByteDance’s most important mission; by July, Liang had issued a company-wide letter for the first time in four years, reshaping ByteDance’s organizational structure and leadership principles, emphasizing the elimination of business silos and enhancing organizational agility around core strategies. (Further reading: Liang Ruibo Tears Down the Old ByteDance)

Looking back today, the letter to all employees about reshaping organizational and leadership principles was the signal for this major transformation.

Within less than a month, the organization was realigned according to the strategy.

It sounds light when said, but consider its weight carefully.

Over the past decade, ByteDance's organizational focus has been solely on TikTok—resources, talent, and monetization have all revolved around TikTok.

Today, a new product has begun to possess the same capabilities: Doubao.

Note that this does not refer to DouBao’s user growth or revenue increase—that’s a product-level matter. The core point is that the company’s organizational structure is now being fully reorganized around DouBao.

The Feishu product team has been integrated into DouBao; the Feishu lead reports to the DouBao lead; Feishu’s GTM has been integrated into Volcano Engine’s B2B ecosystem.

A business unit that had operated independently for nearly ten years was split, restructured, and absorbed by Doubao in a single day.

This is a very strong signal. A product only transcends itself and becomes company strategy when it becomes the organizational center.

Using TikTok as a comparison makes this concept clearer.

Before 2018, TikTok was merely a short-form video app under Bytedance. In 2018, TikTok exploded in the market and became the strategic focus of ByteDance.

After years of business integration, ByteDance has gradually rebuilt its domestic operations around Douyin, incorporating e-commerce and local life services into the Douyin ecosystem, while independently developing TikTok for overseas markets. Douyin has ultimately evolved from a standalone product into the foundational pillar of ByteDance’s domestic business.

Today, the same thing is happening to Doubao.

This year, Doubao officially launched its Professional Edition, marking its transition from a free, consumer-facing entertainment chat tool to a productivity engine with advanced reasoning and complex task-handling capabilities.

With the foundational capabilities of the Professional Edition, Doubao Enterprise Edition naturally integrates into Feishu scenarios, transforming from a simple chatbot into an AI assistant deeply embedded in enterprise workflows. It natively connects with Feishu Docs, Sheets, Meetings, and group chats, complemented by enterprise-grade data isolation, permission controls, and security auditing.

ByteDance is transforming Doubao from a consumer-facing chat tool into a B-end productivity foundation.

Feishu has become the connector in this transformation.

Is this role important? Very much so.

If DouBao were just an API, companies could only call the interface to receive text responses. But if DouBao could directly read corporate Feishu documents, analyze meeting minutes, answer business questions in group chats, and turn the knowledge base into an active assistant, it would be an entirely different level of capability.

Feishu gave DouBao a core asset: real office scenarios.

The biggest challenge for enterprise AI isn't model capability—it's identifying the right use cases and enabling AI to know when and how to intervene in workflows. Feishu has accumulated a decade of data and product experience in office scenarios, which cannot be rebuilt from scratch but must be inherited.

This might be an even better outcome than an independent listing.

Feishu no longer needs to prove itself—its value is now fully embedded within a larger narrative.

Many people are paying attention to Feishu's integration into Doubao, but may have overlooked another development: the establishment of a creativity services platform.

The strategic value of this matter may be greater than product integration.

What was the status of ByteDance's B2B business in the past?

Feishu has one sales team, VolcEngine has another sales team. A single enterprise customer might simultaneously interact with three ByteDance representatives—one selling Feishu, one selling cloud services, and one selling models.

Three sales systems, three pricing models, three types of customer relationships. This isn’t B2B—this is the exact opposite of B2B.

Enterprise customers never buy according to your organizational structure. They care only about solutions, not how you’re internally divided—they just want to know if one person can get the job done.

Today, ByteDance consolidated all enterprise-facing channels into a single point of contact.

Lin Chan, head of sales at Feishu, and Shi Zhijuan, head of strategy and marketing at Feishu, both report to Tan Dai, head of Volc Engine.

One customer, one window, one solution.

This is how you do B2B right.

In the B2B arena, product integration solves what to do, while unified GTM solves how to sell.

In B2B, how to sell is often harder than what to sell. Products can be iterated quickly, but rebuilding sales systems, customer relationships, and delivery processes often takes years.

Choose the byte in one step.

This indicates that Byte's goal is not to optimize B2B, but to reconstruct the underlying logic—no longer organizing sales around products, but around customer needs.

What do customers need? AI. Feishu, VolcEngine, and DouBao are all components that meet this need. These components can be developed separately, but for customers, they must be bundled together for sale.

This is the first time Byte has truly thought like a B2B company.

The concept is simple, but over the past few years, Chinese internet companies doing B2B have rarely truly understood it. Everyone has been organizing sales teams by product lines, forcing customers to adapt to the organizational structure.

Today, Byte has reversed this approach, aligning its organizational structure with customer needs.

In the AI era, what should companies organize themselves around?

The answer given by ByteDance is: Focus on the model, not the product.

Outside the layout:

The biggest change Feishu made today wasn't its integration into Doubao, but that it proved one thing: AI is beginning to restructure a company's organizational framework.

In the past, companies decided products. Today, models are beginning to decide companies.

Who reports to whom, where resources are allocated, and which businesses become central are no longer decided by experience—they are now determined by AI.

Byte is just the first, not the last.