The recruitment platform business is unique—its competitive advantage stems from trust. Job seekers trust that the listed positions are genuine, while employers trust that the resumes are valid. The combination of trust from both sides forms the platform's true pricing power.

In March, following the Spring Festival, many people are busy again. Graduates are revising their resumes for the nth time—the number of college graduates nationwide is expected to reach 12.7 million in 2026. Those considering a job change are contemplating whether to switch industries or even relocate. Corporate HR departments are holding back recruitment needs for new teams, waiting for suitable candidates. Restaurant and factory owners, while pleased with increasing orders after resuming work, worry about labor shortages.

These individuals face different situations, but they all share one common action: opening Kanzhun Limited's app. Amidst this bustling activity, on March 18th, before the U.S. stock market opened, Kanzhun Limited released its financial report for the fourth quarter and full year of 2025, showing steady growth.

The data revealed that the platform's revenue for the fourth quarter of 2025 was 2.08 billion yuan, a 14% year-over-year increase. For the full year of 2025, revenue reached 8.27 billion yuan, up 12.4% compared to the previous year. This raises a question: Why do some people feel that finding a job is difficult, yet recruitment platforms report strong performance? How exactly do these platforms generate revenue?

Those familiar with the online recruitment industry know that Kanzhun Limited primarily earns from the enterprise side, with almost no charges to job seekers. According to the financial report, recruitment revenue from enterprise clients accounted for 8.19 billion yuan in 2025, representing 99% of the total annual revenue. Essentially, the company profits from "matching"—enterprises are willing to pay continuously only when they successfully hire through the platform. Similarly, when a company hires someone, it means a job seeker on the other end has found employment.

Outside the mainstream focus, blue-collar and lower-tier market enterprises have been actively recruiting over the past two years. These markets, beyond the white-collar segment in major cities, have become the growth engine for Kanzhun Limited's performance.

Examining recruitment data, the white-collar segment also showed signs of recovery in 2025. Industries such as internet technology, telecommunications, and semiconductors on the platform experienced significantly higher growth rates compared to 2024, indicating structural improvements. After the 2026 Spring Festival, the average daily number of new job seekers and employers on the platform slightly exceeded the same period last year.

With overall recovery and rapid rebounds in specific sectors, corporate hiring intentions are strengthening year by year, presenting new opportunities for Kanzhun Limited. A detailed breakdown of the company's revenue sources reveals that its growth is built on three solid foundations: an increasingly dense bilateral network, an untapped blue ocean in lower-tier markets, and an AI-enhanced user experience and security system.

The recruitment business is fundamentally about matching. Success in matching relies not on advertising-driven popularity but on genuine user engagement. When "browsing Kanzhun Limited when dissatisfied with work" becomes a self-deprecating phrase among office workers, and "landing a job through Kanzhun Limited" represents a common aspiration, it is clear that the platform has become the most prominent recruitment software.

This prominence stems from the "bilateral network effect," which is difficult to replicate in the short term: more job seekers attract more genuine job postings, which in turn draw more employers; more employers offer diverse opportunities, attracting more job seekers to register. These two flywheels interlock, making it hard to stop once they gain momentum.

The data speaks for itself. As of December 31, 2025, Kanzhun Limited had served over 250 million job seekers and more than 20 million enterprises. QuestMobile data shows that in December 2025, the app's monthly active users grew by 12.2% year-over-year, not only surpassing industry peers in growth rate but also in total user volume—essentially equaling the combined monthly active users of the second, third, and fourth-ranked platforms.

In terms of user demographics and engagement frequency, Kanzhun Limited firmly holds the top position in the industry. QuestMobile data indicates that in December 2025, the platform's average monthly usage frequency and duration far exceeded those of competitors. This suggests that users are not merely opening the app briefly but are actively applying, communicating, and comparing opportunities—a sign of "deep usage" rather than casual browsing.

Moreover, under the same macroeconomic conditions, Kanzhun Limited's growth has outpaced other recruitment platforms. While market trends play a role, product strength is the decisive factor. As the platform's leading position becomes more pronounced, its customer acquisition efficiency has improved. Data shows that in the fourth quarter of 2025, sales and marketing expenses were 389 million yuan, down 9% year-over-year. For the full year of 2025, these expenses totaled 1.69 billion yuan, a decrease of 18%. Despite this, user growth remained robust, with nearly 46 million new users added in 2025. In the fourth quarter, the average monthly active users reached 58 million, a 10.1% increase year-over-year. In other words, the platform is attracting more users with less spending—a signal that the bilateral network effect is now "helping the company save costs."

Understanding how the flywheel turns leads to a more critical question: What "fuel" drives this mechanism? A breakdown of revenue sources shows that Kanzhun Limited's sustained growth over the years relies not on price increases but on a "base expansion" strategy—continuously bringing more enterprises without prior online recruitment habits into the paying user pool.

Financial report data indicates that, in the twelve months ending December 31, 2025, the number of paying enterprise clients reached 6.8 million, an 11.5% year-over-year increase. This figure aligns closely with the platform's quarterly revenue growth, both hovering around the teens percentage-wise. It demonstrates that revenue growth is primarily driven by an increase in paying users rather than higher average spending per client.

The reason Kanzhun Limited's revenue growth remains stable without relying on price hikes lies in the free market economy, where prices are not solely determined by the platform. Just as raising stall fees in a market without attracting more vendors won't boost income, simply increasing prices when businesses struggle and hiring intentions decline would drive clients away rather than raise revenue.

The "new payers" Kanzhun Limited has attracted mainly come from two directions. First, the blue-collar market, which exceeds 400 million people. In 2025, the proportion of revenue from blue-collar segments further increased from 38% in the same period the previous year, becoming the most important structural driver of platform growth. Behind this lies an easily overlooked detail: China's blue-collar workforce is getting younger. Unlike the post-70s and post-80s generations, who relied on acquaintances or job fairs, post-90s and post-00s blue-collar workers are accustomed to searching for jobs on their phones. QuestMobile data shows that in December 2025, users aged 24 and below accounted for 31% of Kanzhun Limited's user base, far exceeding competitors. Many of these young users are blue-collar workers. As blue-collar job seekers join, blue-collar employers naturally follow and begin paying for services.

The second direction is small and micro enterprises in lower-tier markets. In 2025, revenue from small and medium-sized enterprises with fewer than 100 employees exceeded 50% for the first time. In the fourth quarter, revenue from third-tier cities and below approached 25%. Hiring needs for small business owners in county towns are straightforward: fast, lightweight, and affordable—aligning perfectly with Kanzhun Limited's product logic. Compared to traditional online recruitment channels like local job boards, the platform offers simplicity; employers can browse and chat directly with candidates during free moments. In terms of pricing, competition for county-level positions is less intense than in major cities. Many basic job postings are free, while paid listings support monthly subscriptions starting at几十元 or几百元, avoiding annual contract thresholds of tens of thousands of yuan. For small restaurants where the owner handles recruitment alone, Kanzhun Limited may not have the most local candidates or the lowest prices, but it offers the best overall experience.

The inclusion of these "new payers" is clearly reflected in the financial data. In 2025, revenue contributions from KA, medium-sized, and small clients all saw growth. The proportion of revenue from small clients (annual spending of 5,000 yuan or less) increased from 40.6% in 2024 to 42.2%—many of these small clients are likely blue-collar industry bosses and county shop owners new to online recruitment. Meanwhile, the revenue share from KA clients (annual spending of 50,000 yuan or more) also rose. The saying "large ships are hard to turn" applies here; even relatively conservative large enterprises are increasing recruitment budgets, making this indicator another signal of economic recovery.

If network effects and blue-collar/lower-tier markets form the current foundation, then AI represents Kanzhun Limited's bet on the future. During the 2026 spring recruitment season, a graduate opening the app might type "looking for AI-related jobs" in the search bar. Clicking "Ask AI," the AI assistant "Zhi Shanshan" quickly recommends multiple positions with matching reasons. The graduate could then refine their resume through AI dialogue and use the platform's AI mock interview tool to practice after receiving an invitation. Data shows that in the fourth quarter, the user base for "Zhi Shanshan" grew by over 200% sequentially. Meanwhile, the AI mock interview tool expanded its services to students and users with 0-5 years of experience.

Unlike general large models, Kanzhun Limited's built-in AI leverages vast platform data on resumes and job postings, making it more attuned to the recruitment industry and user-friendly. For employers, changes are equally evident. The AI Agent automatically matches candidates based on personalized hiring requirements, with internal tests showing a 25% efficiency improvement for users of this feature. In the fourth quarter of 2025, the average query length for AI deep search (formerly AI quick recruitment) was dozens of times longer than traditional searches. Additionally, 30% of employers using the platform's interview room utilized AI for summarizing interview content, indicating AI's integration beyond mere "information retrieval" into detailed recruitment processes.

However, AI's value on Kanzhun Limited extends beyond efficiency. It also addresses a less visible but critical task—gatekeeping. Every suspicious job posting like "monthly salary 30,000 yuan, no experience required" and private messages disguised as "internal referrals" could be traps. AI is fully integrated into safety measures such as enterprise verification, job posting reviews, and violation detection, enabling faster and more accurate risk identification. Previously, the platform's "2025 Security Governance Report" revealed that 80% of banned accounts were proactively intercepted by AI-supported risk control systems.

These applications are powered by Kanzhun Limited's self-developed recruitment-specific model "Nanbeige." Entering 2026, the model gained attention in open-source communities for its "small parameters, strong performance" approach. The February 2026 release of Nanbeige 4.1-3B, with only 3 billion parameters, demonstrated strong general reasoning, code generation, and deep search capabilities—outperforming models of similar size and competing with larger models in tasks like coding, math, alignment, and tool usage. Public reports indicate that Nanbeige 4.1-3B entered the top three on Hugging Face's global model trend ranking upon release, leading the text model trend chart.

For a recruitment platform, developing its own large model is more than a technical badge. It means that whether recommending jobs to seekers, screening resumes for employers, or identifying ambiguous risky messages, the platform relies on its trained "judgment" rather than generic model generalizations. This vertical technological accumulation is the true foundation for effective AI application.

In conclusion, the recruitment platform business is unique—its competitive advantage stems from trust. Job seekers trust that the listed positions are genuine, while employers trust that the resumes are valid. The combination of trust from both sides forms the platform's true pricing power. Looking back at Kanzhun Limited's financial report, revenue growth stems from an expanding base of paying enterprises, which in turn results from improved matching efficiency. Enhanced matching efficiency arises from deep usage by bilateral users, and deep usage reflects blue-collar workers finding suitable jobs and small town business owners hiring needed staff. The bilateral network enriches the ecosystem, lower-tier expansion broadens boundaries, and AI refines matching accuracy—all three foundations are deepening, and the flywheel has yet to reach its fastest spin. This logic chain is the key to understanding the company: in the recruitment business, every cent earned corresponds to a successful match. If the platform profits, it indicates that workers are faring reasonably well.