Beyond Profit Fluctuations: Has Kingsoft Office's Long-Term Logic Changed?

Original Content by Xinmou · Author | Yao Hui

Last night, Kingsoft Office released its semi-annual earnings forecast, sparking discussions across various communities and industry groups with notably divergent views. Some focused on the significant sequential decline in Q2 net profit, perceiving it as a sign of fundamental issues, while others attributed it to accounting adjustments, maintaining confidence in the solid growth of the core business.

Having closely followed the evolution of the AI-powered office sector for much of the past year, I’ll use this forecast as a springboard to discuss the real operational dynamics behind the numbers and the broader industry transformation underway.

01

The Nature of Profit

When analyzing this earnings report, the first step is to distinguish between different profit metrics; otherwise, it’s easy to be misled by surface-level figures.

The first-half performance highlighted in the announcement is impressive, with estimated net profit attributable to the parent company ranging from RMB 2.316 billion to RMB 2.719 billion, more than doubling year over year. However, breaking it down quarterly, Q2’s net profit attributable to the parent company is projected to be only RMB 121 million to RMB 524 million, representing a significant sequential decline.

This stark contrast is the direct source of market disagreement.

The core reason for this fluctuation lies in a substantial investment gain recognized in Q1.

In the first quarter of this year, Kingsoft Office’s equity investment in Zhipu AI completed its listing on the Hong Kong Stock Exchange. Under current accounting standards, significant changes in the fair value of an investee company following its IPO must be recognized as investment gains in the quarter of listing.

This gain was a one-time accounting revaluation, recognized only in Q1, and will not be repeated in Q2 due to daily stock price fluctuations. The absence of this substantial gain in Q2 naturally led to a sequential decline in reported profit—a normal accounting phenomenon unrelated to the performance of the company’s core business.

An often-overlooked detail is that even non-recurring profit from operations (profit after non-recurring items) does not fully reflect the true profitability of the core business. The classification rules for changes in the fair value of financial assets are complex, and some items may still be included in the non-recurring profit category. Relying solely on this metric can also lead to misjudgments about the company’s operational health.

A more informative metric is the adjusted net profit attributable to the parent company, which excludes non-operating factors such as gains from fund investments and share-based compensation.

For the first half of the year, adjusted net profit is expected to range between RMB 1.000 billion and RMB 1.174 billion, representing year-over-year growth of 18.03% to 38.55%, indicating steady growth in the core business.

For Q2 alone, adjusted net profit attributable to the parent company is projected to be between RMB 420 million and RMB 594 million, still showing positive year-over-year growth, though likely declining sequentially.

However, revenue performance in the same period was not weak. Q2 revenue is expected to range between RMB 1.601 billion and RMB 1.799 billion, up 18.07% to 32.74% year over year, with the revenue midpoint slightly increasing sequentially.

With revenue still rising but adjusted profit declining, the implication is clear: costs have increased significantly. Given the company’s strategic direction and industry-wide trends, this additional cost primarily stems from accelerated investments in AI R&D and computing power.

In summary, the profit fluctuations in this forecast are more a result of investment events and accounting rules than a decline in the core business. The market should focus instead on the quality of revenue growth and the efficiency of cost investments.

02

Growth Transition

Compared to the controversy over profit figures, changes in revenue are more indicative of future trends.

The revenue growth rate of 20.95% to 28.43% in the first half of the year represents a significant acceleration from the single-digit growth in the same period of 2025 and outpaces the industry average.

According to IDC’s previous estimates, the overall growth rate of China’s office suite market in 2026 will be around 14%. Kingsoft Office’s growth rate is nearly double the industry average, with its market share continuing to rise.

More importantly, the drivers of this growth have shifted.

In previous years, Kingsoft Office’s growth relied heavily on two waves of dividend (dividends/opportunities): one was the Legitimate substitution (legalization and replacement) driven by government-backed IT innovation policies, and the other was the渗透 dividend (penetration dividend) from individual users transitioning from perpetual licenses to subscription models.

These two waves supported the company’s rapid growth in recent years, but market concerns persisted: once legalization and penetration reached their limits, where would the next growth driver come from?

AI has provided a new answer. Starting in the second half of 2025, the contribution of AI features to revenue gradually became evident, and by the first half of 2026, AI had become one of the core engines of growth.

Let’s start with individual users. As of the first quarter of this year, WPS Office had nearly 680 million monthly active devices globally, with over 320 million monthly active users on PC. The domestically embedded WPS AI has surpassed 80 million monthly active users, with high daily usage rates, indicating that users are gradually forming habits around AI features.

With the launch of premium AI packages like Lingxi Pro, the average revenue per user (ARPU) for individual subscribers has started to rise. In the first quarter of this year, revenue from individual business grew by 13.8% year over year, rebounding from slower growth in previous quarters—a clear sign that AI-driven monetization is taking effect.

However, overall paid penetration remains low, leaving significant room for future growth, though the pace of conversion will require ongoing observation.

The enterprise segment shows even greater growth potential. WPS 365, the company’s offering for corporate clients, has maintained year-over-year growth exceeding 60% for multiple consecutive quarters. In the first quarter of this year, revenue from WPS 365 reached RMB 244 million, up 60.79% year over year, making it the fastest-growing segment within the B2B business.

During this year’s WAIC, the company introduced WPS Comate as the unified AI gateway for WPS 365, further expanding the value proposition of its enterprise products.

By structuring knowledge, expertise, and business processes into three layers—skills, experts, and applications—WPS Comate addresses the fragmentation and siloed nature of AI applications within enterprises.

This repositioning means the product is no longer just an office tool but an organizational knowledge management and task delivery platform. Correspondingly, ARPU has shifted from the traditional few hundred yuan per account per year to enterprise-level solutions priced at tens of thousands of yuan or more, significantly raising the revenue ceiling.

Overseas markets are also contributing new growth. Public data shows that Kingsoft Office’s overseas paid users grew by over 60% year over year, with revenue from individual users abroad increasing by more than 50%.

In the global office software market, Kingsoft has avoided direct competition with Microsoft in mature markets and instead leveraged AI differentiation to gradually gain traction in emerging markets, forming a second growth curve.

Overall, Kingsoft Office’s growth logic has undergone a transition: from policy-driven legalization dividends to product-driven AI value-add, and from selling tool licenses to offering subscription-based services. The quality of this transition will determine the company’s growth trajectory over the next three to five years.

03

The Competitive Landscape and Kingsoft’s Position

Kingsoft Office’s AI investments are not isolated corporate moves but part of a broader AI-driven transformation across the office software sector. Entering 2026, competition in China’s AI-powered office market has clearly entered a new phase, shifting from early feature comparisons to holistic contests involving product ecosystems, strategic layout (ecosystem positioning), and monetization capabilities.

First, consider Microsoft’s moves. Due to compliance and data security requirements, Microsoft’s global version of Copilot has not officially launched in mainland China. In April this year, 21Vianet, through its Azure China operations, introduced a localized version of Copilot for Microsoft 365 users in China, providing compliant AI agent services.

While still lagging behind the global version in functionality, it remains highly attractive to foreign-invested enterprises and large clients heavily reliant on the Microsoft ecosystem, posing long-term pressure on Kingsoft Office’s premium enterprise market.

More direct competition comes from domestic collaborative office vendors, all of which completed AI business restructuring and strategic upgrades this year.

Alibaba consolidated its enterprise-grade agent products, unifying capabilities from Qianwen Office and DingTalk AI under the DingTalk team to create a unified enterprise AI gateway.

DingTalk itself has amassed a large client base in government, manufacturing, and other sectors, with high-frequency workflow entry points like IM and approvals. By embedding AI capabilities into business processes, its client overlap with Kingsoft Office’s B2B customers is substantial.

Around the same time, ByteDance restructured its AI operations, merging the Feishu product team with the Doubao product team and integrating Feishu’s commercialization team with Volcano Engine.

Post-restructuring, Feishu’s AI capabilities can directly leverage Doubao’s large model technology base, accelerating product iteration. Public data shows that Feishu’s revenue grew by over 100% year over year in the second quarter of this year, with over 90% of new clients also purchasing AI-related products—a strong growth momentum.

Tencent, meanwhile, is rapidly scaling its AI office agent offerings for desktop users.

According to Analysys monitoring data, monthly visits to mainstream desktop-based AI office agent platforms in China exceeded 60 million in June 2026, with Tencent WorkBuddy ranking first, accounting for more visits than the second and third platforms combined. While currently focused on lightweight office scenarios, its potential to expand into deep office workflows is significant, given its access to WeChat and Enterprise WeChat’s user traffic.

In this competitive landscape, Kingsoft Office’s strengths and weaknesses are clear.

With over three decades of experience in office software, Kingsoft maintains unparalleled mastery of native document, spreadsheet, and presentation formats, along with a deep understanding of user office habits. WPS AI’s recent top ranking on the TableBench benchmark, with scores exceeding 90 in factual verification and numerical reasoning, demonstrates its technical edge in vertical scenarios like spreadsheet data processing.

The core of office AI competition always revolves around handling complex business data and delivering usable native files—precisely Kingsoft’s strength.

Its weakness lies in the breadth of entry points. Unlike DingTalk and Feishu, which offer high-frequency features like IM, scheduling, and approvals, WPS is primarily used for document editing, resulting in naturally lower user engagement and session duration.

Today, sector boundaries are blurring. Tool vendors are expanding into collaboration and knowledge management, while collaboration vendors are penetrating document and productivity tools. No one wants to remain confined to their original niche.

Ultimately, competition will hinge on understanding customer needs and delivering product services effectively.

04

Profit Pressure: A Necessary Phase of Transformation

Many express concern over the sequential profit decline in Q2, but within the context of AI transformation across the software industry, this is a typical transitional phenomenon—not unique to Kingsoft.

Traditional office software is a classic high-margin business. Once product development is complete, adding users incurs minimal extra costs, allowing industry leaders to sustain gross margins above 85% for years.

The integration of AI features has fundamentally altered this cost structure.

Higher user engagement with AI features directly increases computing costs, shifting the software business from a predominantly fixed-cost model to one combining fixed and variable costs.

Global software leaders like Microsoft and Adobe are experiencing similar declines in gross margin centers. For Chinese vendors, computing costs represent a larger share of revenue, exerting greater pressure on profits.

This explains the industry consensus that revenue growth should take priority over profit growth for AI application companies at this stage.

Only by expanding user scale and engagement can companies dilute unit computing costs through economies of scale while accumulating more data to refine models, creating a virtuous cycle: more users → lower costs → better products → even more users.

Attempting to protect short-term profits by cutting investments risks falling behind in product capabilities and losing long-term market position.

Kingsoft Office’s unique advantage lies in the substantial investment gains from Zhipu’s IPO, providing strategic financial buffer.

While the industry faces a dilemma between investment and profit, Kingsoft can allocate resources to accelerate R&D and computing investments without distorting its annual profit outlook.

In the fast-evolving AI sector, this financial flexibility creates a valuable time window. The performance gap between leading and smaller vendors may widen significantly during this investment phase.

Of course, investment efficiency remains paramount. The key is whether R&D spending translates into product competitiveness and ultimately drives paid conversions and ARPU growth. While Kingsoft’s investments are already reflected in product leadership and revenue growth, the sustained optimization of return on investment requires longer-term validation.

Another industry variable worth monitoring is the development of on-device large models. As these models improve, more inference tasks can shift from cloud to local execution, significantly reducing computing costs. Given WPS’s massive PC user base, Kingsoft could establish a new cost advantage by leading in scalable on-device AI deployment.

Overall, this semi-annual earnings forecast is neither surprisingly strong nor a signal of fundamental deterioration—it reflects a mature software company’s normal performance during AI transformation. There are growth highlights and short-term pressures from investments, along with industry variables requiring ongoing observation.

While WPS AI already boasts a large user base, individual users’ willingness to pay for AI features remains in the cultivation stage. Many still perceive AI as a novelty tool rather than an indispensable feature.

Additionally, competition in the B2B market is intense, and the enterprise office sector does not follow a winner-takes-all dynamic. Client needs vary widely by size and industry: large enterprises prioritize system integration, data security, and collaboration capabilities, while SMEs focus on cost-effectiveness and tool efficiency.

For enterprise products, balancing commercialization and user experience remains critical. Office tools are high-frequency daily products, and users have low tolerance for ads or complex membership tiers. With AI features added, designing membership levels and pricing strategies becomes even more crucial.

Finally, a large company’s performance often mirrors its industry’s current state—eager for AI-driven growth opportunities yet sensitive to short-term volatility. This duality captures the prevailing sentiment today.

— END —For more, follow us below