China's AI Subsidy War Bought Users, Not Habits

Billions Spent, Users Gone: China's AI Subsidy Experiment Backfires

China's tech giants discovered an expensive truth during the 2026 Lunar New Year: you can buy attention, but you cannot purchase genuine user habits. Over 4.5 billion yuan (US$630 million) in subsidies, red envelopes, and giveaways temporarily flooded AI apps with users. Then came the crash.

Alibaba's Qwen committed 3 billion yuan to the campaign. Tencent's Yuanbao distributed 1 billion yuan in digital red envelopes. Baidu pledged 500 million yuan. ByteDance's Doubao joined the spending frenzy with undisclosed millions.

The results were spectacular, briefly. More than 80% of users abandoned the apps once incentives disappeared, leaving a devastating 30-day retention rate of just 12-13%. The broader AI consumer war in China had hit its first major reality check.

The Great Giveaway: Who Spent What

Alibaba's Qwen reached over 130 million users during the February 6-17 campaign period. Daily active users peaked at 73.5 million on February 7, with users placing nearly 200 million orders through the AI assistant for everything from bubble tea to flight bookings.

More than four million users aged 60 and above completed purchases within two to three voice exchanges with the chatbot. This demographic breakthrough suggested AI interfaces could transcend age barriers when properly incentivised.

"It is becoming a new custom during the Spring Festival for users to buy food and beverages, book flights and hotels, and purchase movie and attraction tickets through Qwen." - Alibaba official statement, February 2026

Tencent's Yuanbao recorded 3.6 billion lucky draws and 1 billion AI-generated items during the holiday period. Nearly half of all red-envelope participants, 49%, came from third-tier cities or smaller, proving the campaigns successfully penetrated beyond China's tech-savvy urban centres.

ByteDance's Doubao hit 78.7 million daily active users on February 7 and recorded 1.9 billion interactions on New Year's Eve alone. The numbers looked impressive until users started leaving en masse.

What the Money Actually Purchased

The campaigns achieved something meaningful: introducing AI-assisted commerce to millions of first-time users. Four million users over 60 completing voice-driven purchases represents a genuine interface breakthrough. The technology works across age groups when motivation is sufficient.

But the 80% drop-off reveals the harsh truth. Most users came for free bubble tea, not revolutionary technology. The subsidies bought trials, not habit formation. This pattern echoes previous Chinese tech subsidy wars in ride-hailing, food delivery, and bike-sharing.

"The budget escalated from 500 million yuan to 3 billion yuan in a matter of days. This is an unsustainable subsidy war that mirrors the cash-burning cycles we've seen across Chinese tech sectors." - Tech industry analyst, February 2026

The playbook remains consistent: spend aggressively to acquire users, hope enough stick around to justify investment, then monetise through adjacent services. The difference with AI apps is that chatbots lack the clear daily utility of transport or food delivery.

Winners, Losers, and Strategic Positioning

Alibaba arguably extracted the most value. Qwen's integration with Taobao and Alipay embeds retained users within a comprehensive commerce ecosystem. Tencent's WeChat red envelope approach carried lower risk by leveraging existing social behaviours.

ByteDance and Baidu spent heavily without equivalent commerce infrastructure to capture long-term value. Their retention challenges will prove more severe as the competition for AI dominance intensifies.

The Retention Crisis Nobody Acknowledges

A 12-13% retention rate after 30 days is dismal by consumer app standards. Successful Chinese consumer apps typically achieve 30-day retention above 25%. The gap suggests most users found no compelling reason to return once free incentives disappeared.

This raises fundamental questions about AI chatbot viability as standalone products. Perhaps the interface needs deeper integration into existing workflows, commerce, messaging, and search functions. Users should never perceive AI as a separate product requiring conscious activation.

• Alibaba's Qwen benefits from direct Taobao and Alipay integration, creating the strongest retention pathway through embedded commerce
• Tencent's WeChat integration embeds Yuanbao in social behaviour, though commerce connections remain weaker
• ByteDance and Baidu face the hardest retention challenge without deep infrastructure integration
• The 49% third-tier city participation proves AI can reach beyond urban audiences, but only with financial incentives
• Voice interface adoption among elderly users suggests untapped market potential with proper onboarding

Beyond the Subsidy Wars

Billion-yuan giveaways cannot continue quarterly. No company sustains this spending indefinitely. Competition now shifts to product development: which AI assistant becomes useful enough for unpaid engagement?

China's approach to consumer AI adoption differs fundamentally from Western markets. US and European adoption stems from curiosity and productivity gains. Chinese adoption relies on cash incentives. The critical question is whether cash-driven trials convert to genuine behaviour change at scale.

The industry faces a moment of reckoning. As our analysis of China's war of a hundred models showed, competition is intensifying beyond simple user acquisition metrics.

Why did Chinese AI apps lose 80% of users so quickly?

Users came primarily for financial incentives rather than genuine product interest. Without free bubble tea and red envelopes, most found no compelling reason to continue using AI chatbots in their daily routines.

Which company benefited most from the subsidy war?

Alibaba likely gained the most lasting value through Qwen's integration with existing commerce platforms. Users who remained are embedded within Taobao and Alipay ecosystems, creating natural retention pathways.

How does this compare to previous Chinese tech subsidy wars?

Similar patterns emerged in ride-hailing and food delivery, but those services addressed clear daily needs. AI chatbots lack the obvious utility that sustained user engagement in previous subsidy battles.

What does this mean for global AI companies?

The retention crisis suggests standalone AI chatbots struggle to maintain user engagement without integration into essential services. Global competitors should focus on embedding AI within existing user workflows rather than creating separate applications.

Will Chinese companies repeat these massive spending campaigns?

Unlikely at this scale. The poor retention rates demonstrate that subsidies generate trials but not sustainable user habits. Future investments will likely focus on product development rather than user acquisition incentives.