‍

On the evening of June 29, ByteDance CEO Liang Rubo sent a company-wide email to employees globally, sparking widespread discussion and debate across the internet.

This move deviated from the usual playbook of major internet companies—no elaborate rhetoric or emotional slogans. Instead, it focused on one core task: targeting management.

The essence of this email can be summed up in three words: taking real action.

Why did this company-wide email go viral?

Liang Rubo's email contained substantial information and openly addressed the challenges ByteDance currently faces.

First, he reiterated the unchanging mission of 'inspiring creativity and enriching lives.' However, the notable addition was his statement: 'From the era of recommendations to the AI era, ByteDance's approach to fulfilling its mission remains consistent—leveraging computation for intelligence and enhancing creativity and experience through intelligence.'

This statement effectively sets the tone for ByteDance's AI strategy, positioning it not as a departure but as a natural extension of its mission.

Even more significant are the updated '10 Leadership Principles.' The last time Liang Rubo revised these principles was in 2022. After four years, this adjustment is sharp and impactful:

The first cut: Eliminating 'complacent management.' New additions include 'pursuing high-impact endeavors' and 'daring to set ambitious goals,' explicitly requiring managers not to focus solely on short-term revenue or monthly data but to look beyond daily operations and target long-term, high-value sectors. In essence, managers who rely on past successes to coast along will no longer survive.

The second cut: Managers must 'go to the frontlines.' 'Maintaining a sense of crisis, adopting an external perspective,' and 'diving deep into the frontlines' are listed as separate items. Henceforth, middle and senior management cannot simply sit in offices reviewing reports and listening to briefings; they must regularly immerse themselves in business, product, and user frontlines.

The email also explicitly states that middle management layers detached from business frontlines and serving only as information conduits will see their organizational value continuously diminished. This is a core signal of ByteDance's push for organizational flattening.

The third cut: Eliminating formalism. The 'Context over Control' philosophy is emphasized, requiring managers to reduce multi-level approvals, repetitive meetings, and ineffective reporting. Many teams previously suffered from 'busywork with no substantive output,' which is now directly cited as a negative management example.

The toughest part is the implementation mechanism: The updated 10 Leadership Principles will be directly incorporated into the core criteria for all managers' promotions and annual evaluations. Managers unable to deliver results or detached from business frontlines will face simultaneous adjustments to their roles and responsibilities.

The reason this company-wide email garnered widespread attention boils down to one core factor: ByteDance is actively confronting 'big company syndrome.'

When a supergiant with over 100,000 employees and revenue exceeding a trillion yuan openly admits internal issues like 'inefficiency,' 'formality,' and 'complacent inertia' in an internal letter—and introduces strict evaluations to correct them—it is newsworthy in itself.

What's happening at ByteDance?

Liang Rubo's email did not emerge in a vacuum. ByteDance is at a crossroads of 'fire and ice.'

First, the 'fire.' In 2025, ByteDance's annual revenue is expected to reach approximately $186 billion (RMB 1.31 trillion), up about 20% year-over-year, surpassing Tencent and Alibaba to rank first among Chinese internet companies.

Douyin's e-commerce GMV for 2025 exceeded RMB 4 trillion, growing over 30% and closing in on Pinduoduo. Overseas revenue increased nearly 50% year-over-year, with the proportion of overseas business revenue rising from 25% in 2024 to over 30%.

But the 'ice' is equally biting.

According to multiple media reports, ByteDance's net profit for 2025 shrank dramatically from $33 billion in 2024 to just over $9 billion, a year-over-year decline exceeding 70%. The profit margin plummeted from 25% to single digits.

Revenue rose by 20%, yet profits fell by 70%. Where did the money go?

The answer is two words: AI.

Insiders reveal that the sharp decline in net profit margins stems from ByteDance's substantial increase in AI business investments during the third and fourth quarters of 2025. In 2026, ByteDance raised its AI capital expenditure plan to RMB 200 billion (approximately $30 billion), an increase of at least 25% from the initial plan.

What does RMB 200 billion represent? It equates to 60% of ByteDance's 2025 profits.

This is ByteDance's current reality: While its revenue scale tops China's internet sector, its profits are being rapidly consumed in the AI 'burning money' race.

Rapid revenue growth coupled with a profit collapse—this stark contrast is the true backdrop of Liang Rubo's company-wide email.

Is ByteDance set to fully monetize its AI tools?

Faced with the dilemma of 'rising revenue but falling profits,' ByteDance's strategy is clear: strategically focus on AI while enabling AI to 'generate its own revenue.'

In June of this year, Liang Rubo made a rare appearance at the Volcano Engine FORCE Conference, officially clarifying ByteDance's strategic focus on AI and setting the 2026 annual keyword as 'scaling new heights,' with the core goal of truly mastering large model technology capabilities. In recent years, ByteDance has continuously narrowed its business scope, prioritizing resource allocation to the AI sector.

However, simply throwing money at it is not enough; it must generate revenue.

Douyin's commercialization has accelerated. On June 24, 2026, Douyin officially launched a professional paid subscription service with three pricing tiers. Some media outlets calculated that with Douyin's 345 million monthly active users, even if only 1% subscribe, it could generate over RMB 228 million in monthly revenue.

Will Douyin fully monetize? Probably not.

Free basic functions for consumers and paid premium features are already an industry consensus. ByteDance's likely strategy is 'capturing consumer users while earning profits from businesses,' with Volcano Engine's MaaS business serving as the true profit engine.

In 2025, Volcano Engine's full-category MaaS revenue reached approximately RMB 1.5 billion, while the 2026 revenue target has been raised to RMB 15 billion—ten times the previous year. Among them, the video large model Seedance 2.0 alone generates over RMB 1 billion in monthly revenue.

ByteDance's strategy is clear: Use Douyin's 345 million monthly active users to capture consumer markets while leveraging Volcano Engine's MaaS services to profit from businesses.

Capturing markets for free and reaping profits through paid services—this classic internet business logic is being replicated in the AI era.

ByteDance's 'Spear' and 'Shield'

When viewed within the broader industry landscape, ByteDance's strengths and weaknesses are equally evident.

Unrivaled traffic moat. Douyin's MAU reached 345 million, surpassing the combined monthly active users of second-place QianWen and third-place DeepSeek. Leveraging ecosystems like Douyin, Toutiao, and Xigua Video, ByteDance has established a formidable traffic barrier.

Revenue scale tops the industry. With RMB 1.31 trillion in revenue, ByteDance surpasses Tencent (RMB 751.7 billion) and Alibaba (RMB 996.3 billion), securing the top spot among Chinese internet companies.

Largest AI investment. ByteDance's 2026 AI capital expenditure of RMB 200 billion exceeds the annual CapEx of any listed cloud provider.

However, its weaknesses are equally critical:

Profits devoured by AI. While Tencent's 2025 net profit reached RMB 224.8 billion (up 16% year-over-year) and Alibaba's 2025 fiscal year net profit hit RMB 126 billion (up 77% year-over-year), ByteDance's profit plummeted from $33 billion to $9 billion.

Late start in B2B business. Tencent Cloud and Alibaba Cloud have already achieved Large scale profitability (scaled profitability), while Volcano Engine's MaaS business generated only RMB 1.5 billion in 2025, a significant gap.

The double-edged sword of going public. Not being listed spares ByteDance from quarterly financial report pressures, allowing unrestricted AI investments. However, it also lacks capital market financing channels and equity incentive tools.

Note: As ByteDance is not a publicly traded company, some data is sourced from public reports or media estimates.

This table reveals a harsh reality: ByteDance has the highest revenue but the thinnest profits. In the AI battle, ByteDance is trading profits for the future.

What's the takeaway?

Returning to Liang Rubo's company-wide email.

Some might dismiss it as merely a 'management rectification.' Based on years of observing ByteDance, this email carries far greater weight than a typical internal adjustment.

First, ByteDance is proactively performing 'self-surgery.' When a company achieves China's top revenue and boasts billions of users, 'big company syndrome'—cumbersome processes, bloated hierarchies, slow decision-making, and detachment from frontlines—becomes inevitable.

ByteDance's choice to proactively expose and rectify these issues at its peak requires courage.

Second, ByteDance is redefining 'management.' The most striking aspect of Liang Rubo's email is transforming 'management' from a position into a verb. Managers are not individuals who 'manage' from offices but must 'dive into frontlines, deliver results, and drive growth.' Generals who cannot fight are unworthy of their rank.

Third, ByteDance is betting on a grander future. The RMB 200 billion AI investment, collapsing profits, and tightened management are all short-term pains to pave the way for long-term competitiveness in the AI era.

Liang Rubo put it bluntly: 'Scaling AI peaks is ByteDance's top priority right now.'

From recommendation algorithms to AI large models, ByteDance is undergoing a difficult but necessary transformation. This email is not a signal of crisis but of clarity.

A company's most dangerous moment is not when it faces difficulties but when it believes it needs no changes. ByteDance is clearly not on that path.

Liang Rubo concluded the email by stating: 'Updating the text is just the beginning; the key lies in implementation.'

Let's see how many managers truly 'dive into frontlines' during the 2026 performance reviews.

That will be the true moment this email takes effect.