Executive Summary / Key Takeaways

• AI-Driven Profitability Inflection: Kanzhun's strategic AI integration across its recruitment platform has triggered a fundamental margin expansion, with Q3 2025 net margins reaching 35.8% while sales and marketing expenses fell 25% year-over-year, proving that artificial intelligence is not just a feature but a structural cost advantage that strengthens network effects.

• Blue-Collar and SME Market Dominance: The company's deliberate focus on underserved segments—blue-collar workers, micro-enterprises with under 20 employees, and Tier 3+ cities—has created a durable moat, with these segments delivering the fastest growth and highest revenue contributions while insulating the business from white-collar cyclicality.

• Network Effect Reinforcement: Despite slashing marketing spend, Kanzhun added 40 million verified users in the first ten months of 2025, demonstrating that AI-enhanced matching accuracy and user experience have made customer acquisition more efficient, creating a self-reinforcing flywheel that competitors cannot easily replicate.

• Capital Allocation Maturation: The board's approval of an $80 million annual dividend and $250 million share repurchase program, funded by RMB 1.2 billion in quarterly operating cash flow, signals management's confidence that the business has transitioned from high-growth cash consumption to sustainable cash generation.

• Critical Risk Monitor: The proliferation of AI-generated content—200 million AI resumes and 40 million AI job descriptions on the platform—threatens the authenticity trust that underpins Kanzhun's blue-collar strategy, requiring constant vigilance to prevent the "massive pollution" that management explicitly warns could undermine user trust.

Setting the Scene: The Recruitment Platform That China’s Underserved Workforce Actually Uses

Kanzhun Limited, founded in 2013 and headquartered in Beijing, operates BOSS Zhipin, China’s largest direct-messaging recruitment platform by monthly active users. Unlike traditional job boards that function as static listing services, BOSS Zhipin facilitates real-time conversations between employers and job seekers, creating a dynamic marketplace where hiring decisions happen through chat-based interactions rather than formal application processes. This architecture reduces friction for the 63.8 million average monthly active users who generated 1.85 billion successful resume exchanges in 2024, making it the most efficient matching engine in China’s fragmented recruitment market.

The company’s strategic positioning defies conventional wisdom. While competitors chased white-collar professionals in Tier 1 cities, Kanzhun built its foundation serving blue-collar workers, micro-enterprises with fewer than 20 employees, and job seekers in lower-tier cities. This wasn’t a temporary market entry strategy—it was a deliberate moat-building exercise. The 2022 launch of the “Cont project” to verify recruiter authenticity and combat false job postings for blue-collar workers represented a costly investment that temporarily suppressed margins but created the trust layer necessary for sustainable growth. That investment is now paying dividends: blue-collar revenue contribution hit a record high in Q3 2025, while manufacturing industries topped sector growth for five consecutive quarters, proving that serving the underserved creates stickier, faster-growing revenue streams than competing for saturated white-collar markets.

Industry structure reinforces Kanzhun’s advantage. China’s recruitment market remains fragmented, with over 40 million small and medium enterprises representing a massive total addressable market that is still largely offline. Kanzhun’s 8.68 million paid enterprise customers represent less than 25% penetration of this segment, leaving substantial room for expansion even as macroeconomic pressures constrain hiring at large corporations. The platform’s mobile-first, direct-messaging model creates network effects that traditional resume-browsing competitors like 51job and Zhaopin cannot easily replicate, as evidenced by Kanzhun’s ability to grow users while reducing marketing spend—a feat none of its peers have achieved.

Technology, Products, and Strategic Differentiation: AI as the New Moat Layer

Kanzhun’s AI integration extends far beyond chatbot window-dressing. The company has deployed artificial intelligence across three distinct layers: job seeker tools (2C), recruiter efficiency (2B), and internal operations. In Q3 2025, the AI job search assistant launched for all users, delivering personalized position recommendations, real-time query responses, and resume optimization guidance. The impact is measurable: interactions per user with the AI assistant showed significant quarter-over-quarter increases, while job seekers completing mock interviews with the AI coaching feature demonstrated improved activity levels and conversion rates. Every AI interaction generates data that improves matching algorithms, creating a data moat that strengthens with scale.

For recruiters, the AI communication assistance feature integrated into commercial products increased the mutual achievement conversion ratio by 7% in Q3 2025. The phased rollout of AI Quick Hiring is already showing promise, with recruiter reading rates steadily increasing as the platform better understands hiring intentions and compares candidates across the entire user base. Enterprise users testing the AI agent reported 25% improvements in recruitment efficiency, while the AI interview feature extended to campus recruiting supports multiple interview rounds with customizable profiles—directly addressing the fresh graduate market where job openings surged double-digits in Q3 2025 even as job-seeking demand declined. These tools transform Kanzhun from a listing service into an intelligent hiring partner, increasing switching costs for enterprise customers.

Internal AI deployment reveals the depth of Kanzhun’s technological commitment. In Q2 2025, 30% of coding at Beijing headquarters was AI-generated, while a newly established R&D department in another city achieved 70% AI-generated code, dramatically accelerating product iteration speed. Security review efficiency improved 30% through AI automation, allowing the company to handle increased user protection workloads without proportional staff increases. Customer service quality inspection and emotion recognition now leverage AI, improving both user satisfaction and employee well-being. These operational gains translate directly to the 25% year-over-year reduction in sales and marketing expenses and the 2.2 percentage point improvement in gross margin to 85.8%—concrete evidence that AI is structurally lowering the cost base.

The proprietary Nanbeige model , integrated with DeepSeek-R1 , distinguishes Kanzhun from competitors relying on generic AI APIs. Management explicitly notes that high-quality vertical-specific data is the bottleneck for recruitment AI, not computing power. With over 225 million cumulative users and billions of interactions, Kanzhun possesses unique behavioral data that improves matching accuracy in ways that new entrants cannot replicate quickly. This data advantage, built over a decade serving underserved segments, represents the true moat—one that becomes more defensible as AI capabilities mature.

Financial Performance & Segment Dynamics: Margin Expansion as Evidence of Moat Strength

Kanzhun’s Q3 2025 results validate the AI-driven thesis. Revenue grew 13.2% year-over-year to RMB 2.16 billion, accelerating from Q2’s 9.7% pace, while GAAP net profit surged 67.2% to RMB 773 million. The 35.8% net margin represents a structural inflection point, up 11.6 percentage points year-over-year and marking six consecutive quarters of sustainable improvement. Critically, this margin expansion occurred while the company added 15 million verified users in Q1 and 40 million in the first ten months of 2025, proving that growth and profitability are no longer trade-offs.

Segment performance reveals the strategic logic behind the moat. Blue-collar revenue growth led all segments in Q3 2025, with manufacturing posting the fastest growth for five consecutive quarters despite tariff concerns. Urban service sectors showed accelerated year-over-year growth in Q2 2025, and recruitment demand rebounded steadily since April. This resilience provides a natural hedge against macro volatility because blue-collar hiring is less correlated with white-collar economic cycles. The segment’s revenue contribution reaching record highs indicates that Kanzhun’s early authenticity investments created a defensible market position that competitors cannot easily penetrate.

Small and micro-enterprise performance underscores the network effect’s power. Companies with fewer than 20 employees contributed nearly 20% of revenue in Q2 2025, representing the highest growth rate among all customer segments. In Q3 2025, revenue from middle-sized and small accounts increased by 2.2 percentage points, while the year-over-year growth rate of new job postings from companies with under 100 employees significantly exceeded the platform’s overall level. SMEs have higher churn risk than large enterprises, yet Kanzhun’s AI-driven matching and direct-messaging model creates stickiness that traditional job boards lack. The fact that over 80% of paid customers are SMEs developed through Kanzhun’s own model—not taken from peers—proves the company is expanding the market rather than stealing share, with ample runway given China’s 40 million SMEs.

Geographic diversification further insulates the business. Revenue contribution from Tier 3 and below cities continued rising in Q3 2025, reaching over 23% in Q1 2025 after a 3 percentage point increase. While Tier 1 cities showed rebounding demand, the steady growth from lower-tier markets reduces dependence on coastal economic cycles and aligns with China’s domestic consumption shift. This geographic balance, combined with segment diversification, makes Kanzhun’s 13.3% year-over-year growth in paid enterprise customers more sustainable than competitors concentrated in volatile white-collar markets.

Cost structure improvements validate the AI investment thesis. Sales and marketing expenses fell 25% year-over-year in Q3 2025, yet user growth remained robust, indicating that brand recognition and network effects have made paid acquisition more efficient. Adjusted sales and marketing expenses excluding sports sponsorship costs still declined 15% year-over-year, proving the efficiency gains are structural rather than one-off. Gross margin expanded 2.2 percentage points year-over-year to 85.8%, while operating leverage drove adjusted operating margins to 41.8%. These metrics demonstrate that Kanzhun’s AI investments are not R&D expenses but capital expenditures that permanently lower the cost curve, creating a competitive advantage that peers relying on manual processes cannot match.

Outlook, Management Guidance, and Execution Risk

Management’s guidance for 2025 reflects confidence in the AI-powered model. The company targets over 35 million newly verified users and RMB 3 billion in non-GAAP operating profit, representing 30% growth over 2024’s RMB 2.3 billion. This profit target is particularly significant because it comes after management explicitly stated they would “secure high-quality bottom line growth first, then hope to actively look for additional upsides.” The fact that they are now pivoting back toward growth—stating they will not “sacrifice our branding growth to achieve this profitability”—indicates the model has reached an inflection point where margins are self-sustaining.

The Q4 2025 revenue guidance of 12.4-13.5% growth suggests acceleration from Q2’s 9.7% but moderation from Q3’s 13.2%, reflecting management’s conservative approach to balancing growth and profitability. This signals discipline rather than desperation; Kanzhun is not chasing growth at the expense of unit economics. The underlying assumptions—continued user growth, expanding market share, and rebounding enterprise demand—appear achievable given the Q3 data showing enterprise-side user growth outpacing job seekers for the first time in three years, indicating improving supply-demand balance.

Execution risks center on AI scaling and competitive response. While Kanzhun’s AI tools show promising early metrics, the 7% conversion improvement and 25% efficiency gains are from limited rollouts. Scaling these benefits across 8.68 million enterprise customers and 63.8 million monthly active users requires robust infrastructure and continuous model training. Management acknowledges they are “cautiously exploring” AI-hosted recruitment services and bulk placement solutions, suggesting they recognize the risk of moving too quickly and compromising the user trust built through the Cont project.

Competitive dynamics warrant monitoring but do not yet threaten the thesis. Management holds a “relatively conservative” view on AI’s near-term impact on the competitive landscape, noting they have not observed “revolutionary or disruptive changes” from peers. While some leading technology companies have expressed interest in entering recruitment, Kanzhun’s two-year head start in AI investment, proprietary Nanbeige model, and accumulated vertical-specific data create meaningful barriers. The open-source nature of models like DeepSeek-R1 levels the playing field for computing power but cannot replicate Kanzhun’s behavioral data moat, which management correctly identifies as the true bottleneck.

Risks and Asymmetries: What Could Break the Thesis

The most material risk is AI-generated content pollution. CEO Jonathan Peng Zhao explicitly warned about “massive pollution of resume or job description,” noting that the platform hosts over 200 million AI-generated resumes and 40 million AI-generated job descriptions. This directly threatens the authenticity promise that underpins the blue-collar strategy. If job seekers lose trust in recruiter authenticity or employers receive fraudulent applications, the network effect could reverse. Management states they are “alert and prepared,” but the risk is existential: authenticity is the moat’s foundation, and AI pollution could erode it faster than detection systems can adapt.

Macroeconomic sensitivity remains a concern despite diversification. While blue-collar and SME segments show resilience, a severe economic downturn would eventually impact hiring volumes across all segments. The company’s low debt (0.01 debt-to-equity) and RMB 19.2 billion cash position provide cushion, but revenue growth could decelerate sharply if enterprise customers reduce recruitment budgets. The tariff war’s limited impact—due to low export-related exposure—demonstrates management’s prudent diversification, but cannot fully insulate against a broad domestic slowdown.

Competitive threats from technology giants could materialize. While management downplays near-term disruption, the observation that “some leading technology companies who have been empowered by AI have expressed their interest in entering the recruitment industry” suggests potential entrants with deeper pockets and existing user ecosystems. If a WeChat or ByteDance leveraged their social graphs to launch recruitment features, they could fragment user attention and pressure Kanzhun’s customer acquisition costs. The company’s current advantage in marketing efficiency might not persist against competitors who can cross-sell recruitment services to billions of existing users at near-zero marginal cost.

Fresh graduate market dynamics present a near-term opportunity but long-term risk. The Q3 2025 data showing double-digit declines in job-seeking demand alongside double-digit increases in job openings suggests a temporary supply-demand imbalance that benefits recruiters. However, if this reflects structural youth unemployment rather than cyclical mismatch, the long-term user growth story could weaken. Kanzhun’s AI interview coaching for recent graduates addresses this, but cannot solve macro labor market dysfunction.

Valuation Context: Premium Pricing for a Compounders’ Quality

At $21.88 per share, Kanzhun trades at a P/E ratio of 28.4, EV/Revenue of 6.6, and EV/EBITDA of 19.7. These multiples command a premium to domestic competitor Liepin (6100.HK), which trades at 9.6x earnings but generates only 6.1% operating margins compared to Kanzhun’s 31.8%. The valuation gap reflects Kanzhun’s superior growth (13.2% revenue growth vs. Liepin’s revenue decline), margin structure (85.8% gross margin vs. 75.8%), and market leadership. For investors, the premium is justified not by current earnings but by the durability of the moat: network effects, AI-driven cost advantages, and penetration of underserved markets create a path to sustained double-digit growth with expanding margins.

Cash flow metrics reinforce the quality thesis. Price-to-operating-cash-flow of 17.1 and price-to-free-cash-flow of 21.5 are reasonable for a company generating 45% year-over-year growth in operating cash flow (RMB 1.2 billion in Q3). The balance sheet’s net cash position and 4.18 current ratio provide strategic flexibility for AI investments, acquisitions, or increased shareholder returns. The recent Hong Kong secondary offering—explicitly for liquidity, not fundraising—demonstrates management’s focus on broadening the shareholder base without diluting existing owners, a sign of capital allocation maturity.

Relative to global recruitment peers, Kanzhun’s valuation appears fair for its growth profile. While 51job and Zhaopin trade at lower multiples, their stagnant user growth and lack of AI integration make them structurally different businesses. Kanzhun’s forward P/E of 20x (implied by management’s profit guidance) versus Recruit Holdings (6098.T)’s 31x suggests potential re-rating upside if the company delivers on its RMB 3 billion operating profit target. The key valuation driver is not current earnings but the sustainability of 30%+ net margins combined with 12-15% revenue growth—a rare combination that justifies a premium multiple.

Conclusion: The Making of a Recruitment Compounders

Kanzhun Limited is transitioning from a high-growth technology platform to an AI-powered compounders, where network effects, operational leverage, and capital efficiency converge to create durable shareholder value. The central thesis rests on three pillars: artificial intelligence that structurally reduces costs while improving matching accuracy, dominant market positions in underserved blue-collar and SME segments that insulate against macro volatility, and a self-reinforcing flywheel where user growth becomes more efficient over time.

The evidence for this thesis is compelling. Q3 2025’s 35.8% net margin, achieved while adding 40 million users and reducing marketing spend by 25%, demonstrates that AI is not merely a product feature but a business model transformation. Blue-collar revenue hitting record contributions and SME segments growing faster than enterprise accounts prove that the early strategy of serving overlooked markets created a defensible moat. The board’s initiation of dividends and expanded buybacks signals management’s confidence that this profitability is sustainable, not cyclical.

What will determine success? First, execution on AI scaling—maintaining the 7% conversion improvements and 25% efficiency gains as features roll out to 8.68 million enterprise customers. Second, competitive defense—preserving the data moat that open-source models cannot replicate while monitoring for disruptive entry by technology giants. Third, macro resilience—continuing to grow through hiring cycles by leveraging segment diversification.

The risks are real but manageable. AI-generated content pollution threatens the authenticity trust that underpins the blue-collar strategy, requiring constant vigilance. Competitive pressure from well-capitalized tech companies could emerge, though Kanzhun’s two-year head start and proprietary data create meaningful barriers. Economic downturns would eventually impact hiring volumes, but the company’s net cash position and diversified segment exposure provide cushion.

For investors, Kanzhun offers a rare combination: a leading market position in a massive underpenetrated market, AI-driven margin expansion, and maturing capital allocation. At 28x earnings, the stock prices in execution but not perfection. The key variable is whether management can scale AI benefits while preserving the user trust built through years of authenticity investments. If they succeed, Kanzhun will not just be China’s largest recruitment platform—it will be one of its most profitable network-effect compounders.