Latest / The CEO Diary with Fexingo: Leadership Lessons, Executive Decisions, and Corner Office Stories / How Jin Li turned Alibaba into a cloud giant
Transcript
- Lucas: When Alibaba Cloud first launched in 2009, it was basically a side project — a way for Alibaba to handle its own e-commerce traffic spikes during Singles' Day. No one outside the company took it seriously as a cloud provider. Luna: And now? It's the largest cloud provider in China and the fourth-largest globally, behind AWS, Azure, and Google Cloud. Lucas: Right. And the person responsible for that transformation is Jin Li — who took over Alibaba Cloud in 2014 and turned it into a $10-billion-plus business. What's interesting is how he did it, because the playbook was very different from what AWS or Azure did. Luna: Different how? I mean, at the core, cloud is cloud — you rent compute and storage. Lucas: True, but the market dynamics are completely different. In the US, AWS had a first-mover advantage and basically defined the market. In China, Alibaba was competing against Tencent, Baidu, Huawei, and a bunch of state-backed players. Jin Li's bet was that the only way to win was to own the entire stack — from the data center hardware up to the application layer. Luna: So not just renting servers, but building them. Lucas: Exactly. In 2013, Alibaba made a huge bet: they decided to build their own data center infrastructure, including custom servers, networking gear, and even cooling systems. That was controversial internally because it required massive upfront capital. But Jin Li argued that if they didn't control the hardware, they'd always be at the mercy of suppliers like Intel and Cisco. Luna: And that bet paid off during Singles' Day, when they needed to scale to handle millions of transactions per second. Lucas: It did. By 2015, Alibaba Cloud was handling 80 percent of Alibaba's own traffic. That gave Jin Li a powerful proof point: if it works for the world's biggest e-commerce event, it can work for any enterprise. He then pushed the cloud team to build services specifically for Chinese enterprises — things like financial compliance tools for banks, and data localization features for government contracts. Luna: That localization angle is huge. Chinese companies have very different regulatory requirements compared to US firms. Lucas: Right. And Jin Li was smart about it. He didn't try to copy AWS's catalog of 200-plus services. Instead, Alibaba Cloud focused on maybe 50 core services that were most relevant to Chinese customers — like database services optimized for Chinese character sets, and AI services trained on Chinese data. That focus allowed them to iterate faster than Tencent or Huawei. Luna: But wasn't there a brutal price war? I remember around 2017, Alibaba started cutting prices aggressively. Lucas: Yes, and that was another Jin Li decision. He launched what they called the 'Cloud Price War' — slashing prices by 30 to 50 percent on core services like compute and storage. The idea was to starve competitors of revenue while Alibaba subsidized the cloud business with profits from e-commerce. A lot of small cloud providers just couldn't survive. Luna: That sounds like classic market-share-over-profitability strategy. Did it work? Lucas: It did. By 2019, Alibaba Cloud had over 40 percent of the Chinese market, with Tencent at around 20 percent and everyone else below 10. Jin Li had effectively consolidated the market. And then he started raising prices — not across the board, but on premium services like AI and database management, where margins were higher. Luna: So the classic 'loss leader then harvest' move. Lucas: Exactly. But what's more interesting is what happened next. In 2020, Alibaba Cloud became profitable for the first time — and Jin Li announced that the cloud division would be run as a standalone business, with its own P&L. That changed the incentive structure completely. Luna: What do you mean? Lucas: Before 2020, the cloud team was essentially a cost center supporting Alibaba's e-commerce. After the restructuring, Jin Li made every product manager responsible for their own unit economics. If a service wasn't profitable, you had to either cut costs or kill it. That forced a lot of discipline. Luna: And how did the culture shift? I imagine a lot of engineers didn't love being told to think about margins. Lucas: That's exactly the tension. Jin Li is known for being extremely data-driven — he once said, 'If you can't measure it, you can't manage it.' He introduced weekly revenue reviews where PMs had to explain any deviation from forecast. But he also invested heavily in R&D — Alibaba Cloud's R&D spending as a percentage of revenue is actually higher than AWS's, around 15 percent. Luna: So he's pushing profitability but also betting on the future. That's a hard balance. Lucas: It is. And the biggest future bet is AI. In 2023, Alibaba Cloud launched its own large language model called Tongyi Qianwen, and Jin Li committed to building out massive AI infrastructure. He pre-ordered NVIDIA's H100 chips in huge volumes — some reports say Alibaba had the single largest order of H100s in Asia outside of the hyperscalers. Luna: That's a huge capital commitment. Are they building those chips into their own data centers? Lucas: Yes. Alibaba Cloud now offers what they call 'elastic GPU clusters' — basically, you can rent NVIDIA H100s by the hour. And they're building data centers specifically optimized for AI workloads, with liquid cooling and high-bandwidth interconnects. Jin Li's view is that AI will be the killer app for cloud — that enterprises will move to the cloud not just for cost savings, but for access to AI compute. Luna: That's the same thesis AWS and Azure are betting on. But in China, the regulatory environment is different — there's more control over data and AI models. Lucas: Exactly. And that's where Jin Li's strategy of being deeply integrated with the Chinese government pays off. Alibaba Cloud is the official cloud provider for several provincial governments, and they've built AI services that comply with China's strict data sovereignty laws. That's a moat that AWS can't easily cross. Luna: So what's the biggest challenge facing Jin Li now, as of mid-2026? Lucas: Two things. One is the us china chip war. The export restrictions on advanced semiconductors mean Alibaba can't get the latest NVIDIA chips — they're limited to the H100, not the B200 or whatever comes next. Jin Li has been investing in homegrown AI chips through Alibaba's semiconductor arm, but that'll take years. Luna: And the second? Lucas: The second is competition from Huawei Cloud. Huawei has been aggressively building out its cloud business, especially for government and state-owned enterprises. They have a reputation for security and reliability that Alibaba sometimes struggles with, especially after the 2023 outage that took down a bunch of services for hours. Luna: So Jin Li is fighting on two fronts — technology access and trust. Lucas: Right. But if you look at his track record, he's been able to navigate these challenges before. He took a division that was losing money and turned it into a cash cow. He outmaneuvered Tencent and Baidu in the cloud wars. And now he's positioning Alibaba Cloud as the AI infrastructure provider for China. Luna: It's a fascinating case study in long-term strategic thinking. Most cloud leaders in the West don't have to deal with that level of geopolitical complexity. Lucas: Exactly. Jin Li's story is a reminder that leadership isn't just about making the right product decisions — it's about understanding the political and regulatory environment you operate in, and building a business that can survive those constraints. Luna: You know, listening to you lay out all that — the pricing war, the infrastructure bet, the AI pivot — it's a lot of decisions that had to be made with conviction. It's the kind of deep dive that makes you think differently about a company you thought you already understood. Lucas: I think that's the value of these conversations. If they help you see one strategic move more clearly, or question an assumption you had about a market, that's worth something. Luna: And if that something is worth, say, the price of a coffee, there's a way to keep these episodes coming — no ads, just listener support. Lucas: Yeah. If today's episode moved your thinking forward in some small way, there's a link at buy me a coffee dot com slash fexingo. It's the smallest possible ask, and it genuinely helps us keep doing this work. Luna: Alright. Back to Jin Li. One thing I've always wondered: how did someone with a background in engineering — not business — end up being such a sharp strategist? Lucas: That's a good question. Jin Li studied computer science at Zhejiang University, then spent years at Alibaba building the company's internal tech infrastructure. He wasn't a typical mba type leader. But what he had was a deep understanding of what the technology could do, combined with a willingness to learn the business side. He once said, 'I don't need to know finance as well as a CFO — I need to know enough to ask the right questions.' Luna: That's a great philosophy for any leader, really. Lucas: It is. And it's why Alibaba Cloud has survived where others have failed. Jin Li built a culture where technical excellence and commercial discipline are not in conflict — they reinforce each other. Luna: So what's the one thing you'd take away from his story? Lucas: For me, it's that big bets require patience. Jin Li didn't turn Alibaba Cloud around in a year. He spent nearly a decade iterating, adjusting, and staying the course. Most leaders would have given up or pivoted too early. He didn't. Luna: And now it's a business that generates billions in revenue and is central to China's AI ambitions. That's the payoff. Lucas: Exactly. Jin Li's story is still being written, but the first few chapters are already a masterclass in strategic leadership under constraints.