Alibaba just reported a quarter that looks terrible on the surface — revenue missed estimates, net income dropped 66%, and operating income collapsed 74%. But the story isn't about a company failing. It's about a company deliberately burning money to win the AI race. Alibaba is spending aggressively on cloud infrastructure, building its own GPU chips, and scaling its Qwen AI models to over a billion downloads. The question is whether that bet pays off — or whether investors are just watching profits disappear.
Here's what happened.
The Numbers: Revenue Missed, Profits Collapsed
- Revenue: RMB 284.8 billion ($40.7B), up only 2% year-over-year — missed estimate of RMB 290.7B
- Adjusted Revenue: Up 9% excluding Sun Art and Intime divestitures
- Net Income: RMB 15.6 billion, down 66% from RMB 46.4B a year ago
- Operating Income: RMB 10.6 billion, down 74% year-over-year
- Non-GAAP EPS: RMB 7.09 ($1.01) per ADS
- Cloud Revenue: Up 36% year-over-year — the standout number
- Free Cash Flow: Down 71% to RMB 11.3 billion
- Cash Position: RMB 560 billion ($80.1B) — still a fortress balance sheet
Translation: Alibaba's headline numbers look bad — revenue missed, profits dropped, and free cash flow dried up. But there are two layers to this story. First, the company recently divested (sold off) Sun Art and Intime, two big retail businesses. If you strip those out, revenue actually grew 9%, not 2%. Second, the profit collapse is mostly intentional — Alibaba is pouring money into AI investments, quick commerce (fast delivery), and user experience improvements. The question is whether that spending will generate returns, or whether it's just money down the drain.
Cloud Intelligence: The Growth Engine
The headline number investors were watching was cloud revenue, and Alibaba delivered: 36% growth year-over-year. That's a meaningful acceleration from previous quarters and puts Alibaba's cloud business in the same growth conversation as the US hyperscalers.
The growth is being driven by AI demand. Alibaba's cloud infrastructure is powering a growing number of AI workloads, both from external customers and internally. The company's Qwen AI models surpassed 1 billion cumulative downloads on Hugging Face, making it the most widely used open-source AI model family in the world.
Perhaps more significantly, Alibaba's chip subsidiary, T-Head, has brought its proprietary GPU into production at scale. This means Alibaba is building its own AI chips — reducing its dependence on Nvidia and lowering its long-term infrastructure costs. The chips are already supporting training, fine-tuning, and inference workloads across Alibaba's cloud platform.
Translation: "Open-source" means the AI models are freely available for anyone to download and use, unlike proprietary models from OpenAI. This is a strategic choice — by making Qwen free, Alibaba builds an ecosystem where millions of developers build on top of their technology. That creates demand for Alibaba's cloud services (where the models actually run), which is how they make money. Think of it like giving away the razor to sell the blades.
Like these translations?
Reading the full earnings report yourself? The Ask AYO extension highlights and translates the jargon in real time — so you can read any company's press release, 10-K filing, or investor call transcript and actually understand it. Free.
Get the free extensionQwen App: Alibaba's ChatGPT Moment
Alibaba's consumer-facing AI application, the Qwen app, has now surpassed 300 million users. Powered by the Qwen 3.5 foundation model, it represents Alibaba's attempt to turn its AI research into a product that everyday people use.
In January 2026, Alibaba integrated Taobao Instant Commerce into the Qwen app, meaning users can now order products for fast delivery directly from within an AI conversation. This is a significant strategic move — it connects Alibaba's commerce engine to its AI platform, creating a loop where AI drives shopping and shopping drives AI usage.
Translation: Imagine asking ChatGPT to recommend a birthday present, and it not only suggests something but lets you buy it and have it delivered in 30 minutes — all within the same app. That's what Alibaba is building. The AI becomes the storefront. It's the kind of integration that Western tech companies are still figuring out, but which makes intuitive sense for a company that owns both the AI and the commerce platform.
Taobao and Tmall: The Core Commerce Story
Alibaba's domestic commerce business (Taobao and Tmall Group) generated RMB 107.1 billion ($14.7B) in revenue this quarter. Growth slowed compared to previous quarters, which the company attributed to weaker transaction activity and the phasing out of software service fees it introduced in 2024.
The bright spots: 88VIP members surpassed 59 million, growing double digits year-over-year. This is Alibaba's premium membership programme (think Amazon Prime for China), and the continued growth shows that the most valuable customers are getting more engaged, not less.
Monthly active consumers on the Taobao app also grew double digits, driven primarily by the quick commerce business — Alibaba's push into ultra-fast delivery (30 minutes or less). This is where a lot of the investment spending is going.
Translation: "Quick commerce" is the race to deliver products in under an hour — sometimes in 15 minutes. In China, this is becoming the standard expectation for online shopping, not the exception. Alibaba is investing heavily in the warehouses, delivery networks, and logistics required to make this work at scale. It's expensive, which is one of the main reasons operating income dropped 74%. The bet is that once the infrastructure is built, the unit economics improve and the investment pays for itself.
International: Losses Narrowing
Alibaba's international commerce division (AIDC), which includes Lazada, AliExpress, and Trendyol, narrowed its losses significantly this quarter. The improvement came from better cost efficiency and growing scale in Southeast Asia, Europe, and Turkey.
This matters because international expansion has been a persistent cash drain for Alibaba. The fact that losses are shrinking, not growing, suggests the overseas business may be approaching a tipping point where it starts generating returns instead of consuming them.
The Spending Decision: Why Profits Collapsed
Let's be direct about why operating income dropped 74%. It wasn't because Alibaba's business is failing. It was a deliberate capital allocation decision. The company chose to invest in three expensive areas simultaneously: AI infrastructure (cloud, chips, Qwen models), quick commerce logistics, and user experience improvements across its platforms.
Management framed this as "investing ahead of the curve" — spending now to build capabilities that will generate revenue in 2027 and beyond. The $80 billion cash pile gives them the runway to do this without needing to raise money. But investors don't tend to like watching profits disappear, even when the spending is strategic.
Translation: "Capital allocation" is how a company decides to spend its money. There are basically three choices: invest in the business (what Alibaba is doing), return cash to shareholders through dividends or buybacks, or save it. Alibaba is choosing option one aggressively. The risk is that the investments don't pay off. The upside is that if AI-driven commerce works, Alibaba could be years ahead of competitors. Amazon made the same bet on AWS in 2006 — losing money for years before it became the most profitable business in tech.
The Bottom Line for Investors
Alibaba missed revenue estimates and saw profits collapse. But this is an investment story, not a decline story. The company is deliberately spending on AI (Qwen hit 1 billion downloads, custom GPUs in production), quick commerce (30-minute delivery at scale), and cloud infrastructure (36% growth). The $80 billion cash position gives them years of runway. The question is whether investors have the patience to wait for the payoff.
↑ Why This Matters (Bull Case)
Alibaba's cloud business grew 36% and is becoming the backbone of AI in China. Qwen is the world's most used open-source AI model family. The company is building its own GPU chips, reducing Nvidia dependence. Quick commerce is driving double-digit user growth on Taobao. International losses are narrowing. 88VIP membership passed 59 million and growing. The stock trades at roughly 10x earnings — cheap by any measure for a tech company investing in AI at this scale. And with $80 billion in cash, Alibaba can afford to keep spending without raising capital.
↓ Why This Might Worry You (Bear Case)
Revenue missed estimates and only grew 2% on a reported basis. Net income dropped 66%. Operating income dropped 74%. Free cash flow dropped 71%. These aren't small numbers. Quick commerce is expensive, and there's no guarantee it will ever be profitable at scale. China's consumer spending environment remains cautious. The company is divesting mature retail businesses while spending on unproven ones. And every quarter of declining profits erodes confidence that the AI investment thesis will actually materialise. Management says this is strategic, but from a shareholder perspective, it looks like a company trading profitability for growth that hasn't arrived yet.
The question is whether Alibaba's AI bet will follow the Amazon playbook — years of investment followed by enormous returns — or whether it's burning through profits in a market where the payoff never comes.
References:
- Alibaba Group — December Quarter 2025 Earnings Press Release (March 19, 2026)
- CNBC — Alibaba Revenue Misses Estimates as Net Income Drops 66% (March 19, 2026)
- Grafa — Alibaba Q3 2025 Earnings: Cloud Growth, Quick Commerce (March 19, 2026)
Ticker: BABA (NYSE) / 9988 (HKEX) · Reported: March 19, 2026
The biggest earnings, translated. Weekly.
One email a week covering what the brands you care about actually said — in plain English. No jargon, no fluff, no spam.
Unsubscribe anytime.