Concerns regarding the consumer-side AI monetization potential of Chinese internet giants may be a misjudgment by the market. According to a recent report from HSBC released on April 2, the AI monetization capabilities of Alibaba and Tencent are systematically undervalued by the market. The potential for AI-driven revenue in core scenarios such as advertising, e-commerce commissions, and payments has not been fully reflected in their stock prices.
Despite Alibaba and Tencent recently disclosing large-scale AI investment plans and raising their cloud business revenue forecasts—with Alibaba setting a target to achieve $100 billion in cloud revenue within five years, implying a compound annual growth rate of over 40%—both companies' stock prices experienced a pullback following their earnings announcements. This is not a signal of market pessimism towards AI prospects but rather reflects short-term investor concerns about rising operating expense ratios due to consumer-facing AI investments.
If Tencent, Alibaba, and ByteDance can capture 10% to 50% of the advertising revenue share from transaction and search-based platforms, with the gains divided equally among the three, this opportunity could provide up to approximately 8% to 11% upside to Tencent and Alibaba's 2027 revenues, respectively. Based on this analysis, HSBC maintains a Buy rating on Alibaba and Tencent, with target prices of $180 and HK$750, representing potential upside of about 44% and 51% from current price levels.
The sell-off in Alibaba and Tencent shares stems from short-term worries about rising operating expenses, not doubts about long-term AI returns. The report suggests that the market remains confident in the long-term return on capital expenditure and the sustainability of free cash flow and balance sheets for both companies. The primary concern is the increase in operating expense ratios driven by consumer AI investments.
The largest monetization opportunity for consumer AI lies in advertising. The report forecasts that China's online advertising market will reach RMB 1.91 trillion by 2027, based on growth rates of 10% in 2026 and 9% in 2027. It is estimated that transaction and search platforms (excluding Tencent, Alibaba, and ByteDance) will hold approximately 30% of the advertising market share by then.
If Tencent, Alibaba, and ByteDance leverage their larger user bases, higher usage frequency, and stronger closed-loop transaction capabilities to capture 10% to 50% of advertising revenue from these platforms, with gains shared equally, Tencent's total revenue could see an upside of 2.1% to 10.6%, while Alibaba's could increase by 1.5% to 7.5%.
Should AI expand overall consumer demand, increase online penetration rates, or if platforms adopt subscription-based收费models following overseas peers, these estimates could see further upside.
Tencent's advantage lies in the high stickiness and diverse monetization paths within its WeChat ecosystem. Data shows that WeChat's DAU/MAU ratio reached 82% in February 2026, significantly higher than Doubao's 30% and Qwen's 16%. Users average 40 sessions per day, far exceeding the 4 to 5 sessions for Doubao and Qwen. This high-quality, high-frequency user base creates a formidable network effect moat.
For monetization, Tencent is deeply integrating OpenAI-like AI product features into WeChat, covering scenarios such as chatting, search, Video Channels, Official Accounts, Mini Programs, and Small Stores to build next-generation AI agent services. Tencent's revenue upside will come from three areas: advertising—improving advertiser ROI through AI automation in targeting, bidding, and placement, thereby increasing conversion rates under equivalent ad loads; commissions—AI-driven shopping experiences boosting Small Store GMV; and payment fees—increased transaction volume from Mini Programs and Small Stores directly contributing to payment revenue.
On the enterprise side, Tencent Cloud achieved an adjusted operating profit of RMB 5 billion in 2025, with revenue growth rebounding. Upside for Tencent Cloud will come from PaaS and SaaS layers, including proprietary AI agent products like WorkBuddy and Qclaw, as well as AI enhancements for collaboration tools such as WeCom and Tencent Meeting.
Alibaba has clear AI monetization paths in both enterprise and consumer segments. On the enterprise side, Alibaba Cloud's external cloud revenue grew 35% year-over-year in the December 2025 quarter, with AI-related revenue maintaining triple-digit growth for ten consecutive quarters. Alibaba leads in China's LLM token market share, holding 32% in the second half of 2025, with platform token consumption growing sixfold over the past three months. Its in-house chip division, T-Head, had cumulatively shipped 470,000 AI chips by February 2026, with annual revenue exceeding RMB 10 billion. Over 60% of T-Head chips serve external customers, covering more than 400 corporate clients in sectors like internet, financial services, and autonomous driving.
On the consumer side, Alibaba's Qwen consumer monthly active users reached 300 million in February 2026, integrated across platforms including Taobao Tmall, Alipay, Fliggy, Damai, and Amap. By the end of February 2026, 140 million users had completed their first AI-driven shopping experience via Qwen's agent features, covering scenarios like food delivery, fresh groceries, ticketing, and travel bookings. As user scale, frequency, and stickiness increase further, this will lay the foundation for monetizing commissions and advertising revenue in transaction-related verticals.
The report clearly expresses a preference for platform-based internet companies, viewing them as having structural advantages over pure-model companies. Within China's foundational model ecosystem, participants are divided into four layers: chips, infrastructure, models, and applications. Alibaba is the only company covering all four layers, while ByteDance is accelerating its in-house chip development. In contrast, pure-model companies like Zhipu and MiniMax rely on cloud service providers for computing power, making their revenue and costs vulnerable if CSPs prioritize internal use or raise prices.
HSBC assigns Hold ratings to both Zhipu and MiniMax, with target prices of HK$920 and HK$1,000, respectively. Both stocks have surged 542% and 687% year-to-date since their listings (while the Hang Seng Index fell 1%), with valuations largely reflecting their model capabilities. Computing power bottlenecks may constrain further revenue growth upside.
In terms of enterprise competitive landscape—assessed across AI computing power (GPU reserves, cloud ecosystem, in-house chips), MaaS (token market share, Artificial Analysis Intelligence Index ranking, modality coverage), and coding capability—Alibaba and ByteDance are in a stronger position for enterprise monetization, while Zhipu and MiniMax lead in coding capability.
Comments