Gunjo · Business Intelligence for the AI Era
← Sticker Wall JOURNEY · DETAIL

Alibaba Cloud: From E-commerce Internal Infrastructure to China's Largest Public Cloud Service Provider

Founded: Wang Jian · Alibaba Cloud

JOURNEY

Key Fields

FIELD STAMPS
IndustryConglomerate / Trading House
RegionChina
ScaleGiant
ChannelOther

Origin

Around 2008, Alibaba's e-commerce and financial transaction scale expanded rapidly, and traditional IOE architectures struggled to support Singles' Day peaks and massive data volumes. Self-developed cloud computing became a survival necessity for the group rather than an optional layout. Wang Jian drove the independent research and development of the Apsara operating system and cloud servers from scratch, aiming not only to serve internal Taobao but also to allow external SMEs to use pay-as-you-go computing power. This starting point meant Alibaba Cloud was long bound to the group's traffic, but it also sowed the seeds for later independent competition through the tug-of-war between parent dependence and market-driven paths.

Milestones

2010
Internal Technology Incubation Period Turning Point
In 2009, Alibaba Cloud was officially established, positioned as China's first independently developed cloud service provider. However, the market generally doubted that an e-commerce company could build an operating system-level cloud platform. In 2010, Apsara cloud servers and ECS were launched, publicly selling the pay-as-you-go model for the first time. However, external customers were sparse, relying primarily on internal group businesses like Taobao and 1688. This stage lacked an independent revenue caliber, the team was long branded as a 'fraudulent project,' external willingness to pay was extremely low, and the commercialization path was once blurred.
2013
From Internal to Limited External Commercialization Failure
In 2013, Alibaba Cloud attempted to sell cloud resources to small and medium-sized e-commerce and government clients, but insufficient stability and poor reputation—especially large-scale availability issues around 2012—severely damaged customer trust. At the time, China's public cloud market was dominated by foreign capital and IDC vendors, with high price wars and customer education costs. Alibaba Cloud's external revenue volume was far smaller than the group's investment, and it was once viewed internally as a subsidiary department that only burned money without independent survival capability.
2017
Scale Commercialization and Government-Enterprise Breakthrough PMF
Starting in 2014, leveraging the spillover effects of Alibaba Group's e-commerce ecosystem and cloud service price-cutting strategies, Alibaba Cloud gradually expanded its base of SMEs and internet company clients. In 2015, Wang Jian stepped down as president of Alibaba Cloud but continued to guide its technical direction. Concurrently, Alibaba Cloud began large-scale recruitment of government and enterprise sales teams, expanding from internet clients to traditional industries and government markets. By fiscal year 2017, Alibaba Cloud's operating revenue reached approximately 6.663 billion RMB, maintaining triple-digit growth for multiple consecutive quarters, marking the first time its financial reports reflected a commercial scale close to that of an independent cloud vendor.
2021
Full Localization and Market Share Leadership Growth
Alibaba Cloud completed the replacement of foreign databases like Oracle with self-developed distributed databases such as OceanBase and PolarDB in its core Singles' Day trading system, reducing its reliance on foreign capital. Fiscal year 2019 revenue was about 24.702 billion RMB, breaking 40 billion RMB in fiscal year 2020, and reaching 60.12 billion RMB in fiscal year 2021, continuously holding the number one market share in China's public cloud market. However, profit margins remained low for a long time, overseas expansion progressed slowly due to geopolitical influences, and revenue was highly dependent on the Chinese market.
2023
Growth Slowdown and Organizational Spin-off Failure
After fiscal year 2022, Alibaba Cloud's revenue growth rate declined rapidly from highs, particularly in the second quarter of 2022 when growth dropped to just 4%, hitting a record low since listing and exposing issues such as slow collection of payments from government and enterprise clients and spending cuts by internet industry clients. In May 2023, Zhang Yong issued an all-staff letter announcing that Alibaba Cloud would spin off and list, positioning it as a second venture, but in November of the same year, the spin-off plan was canceled, announcing a return to the group and focusing on 'AI-driven, public cloud first.' Such repeated spin-offs caused turbulence in strategic and market expectations, leading many corporate clients to harbor doubts about long-term cooperation stability.
2026
AI Cloud Revaluation and Commercialization Acceleration Turning Point
Starting in 2023, Alibaba Cloud integrated Tongyi Lab, Pingtouge, and the cloud business into the 'Tong-Yun-Ge' system, shifting toward computing power plus model output. By fiscal year 2026, external commercialization revenue growth reached 45%, hitting a 22-quarter high, with AI-related products showing triple-digit growth for 12 consecutive quarters. The CUBE 5.0 modular data center architecture shortened delivery times to 100 days and reduced costs by over 10%, supporting more than 50,000 customers in Hong Kong and a doubled global production capacity. However, the group's adjusted EBITA loss for the AI application segment widened, and massive capital expenditure on cloud infrastructure led to a net outflow of free cash flow, indicating that rapid expansion still comes with profit pressure.

Turning Points

  • In 2010, ECS was launched with sparse external customers, forcing the company to rely on internal group businesses to support technical iteration first, establishing an inside-out survival path.
  • After 2015, shifting from a technical service department to large-scale government-enterprise sales and price wars allowed Alibaba Cloud to truly become a commercial company with an external customer structure.
  • Canceling the spin-off listing in 2023 and shifting to AI-driven and public cloud first marked a return from an independent IPO narrative to the positioning of the group's computing power foundation.
  • In 2026, the CUBE architecture and Token Hub shifted Alibaba Cloud from resource sales to AI application delivery, with computing power and model revenue beginning to serve as the main growth engines.

Failures & Pitfalls

  • Availability incidents and early technical instability around 2012 caused massive churn among early external clients, labeling cloud services as unreliable in the market for several years.
  • Government and enterprise client expansion brought slow payment collections and numerous customized projects; post-fiscal year 2022, revenue growth plunged to single digits, and the independent growth narrative clashed with reality.
  • The 2023 spin-off listing plan was announced and then canceled, causing enterprise clients and capital markets to develop long-term doubts about Alibaba Cloud's governance stability.
  • Continuous investment in AI commercialization led to a net outflow of free cash flow for the group, showing that even with high revenue growth, overall financial performance could still be dragged down.

关键成功要素

  • Using the group's own e-commerce business as an early grindstone, leveraging Singles' Day scenarios to force technical maturity before exporting to external clients.
  • Self-developing the Apsara operating system plus database replacements for the IOE architecture, forming a rare domestic full-stack independent and controllable capability with high barriers in government and domestic substitution markets.
  • Combining price wars and scale effects, using the e-commerce ecosystem's spilled-over customers to dilute marginal costs and suppress second-tier cloud vendors.
  • Shifting to AI plus token monetization in 2026, transforming the cloud from resource wholesaling to model and application layer revenue sources, moving away from pure computing power price competition.

Lessons

  • If an internal platform wants to be independent, it must undergo identity reconstruction from a cost center to a profit center; otherwise, it will forever be treated as a money-burning department.
  • Nurturing technology with a single group client is fine, but if external clients are not integrated quickly enough, the growth ceiling will arrive faster than expected.
  • The government and enterprise market brings revenue but drags down cash flow and standardization levels, and cannot be blindly copied using the internet company model.
  • Strategic wobbling in large technical platforms damages customer trust more than technical defects, with repeated spin-off listings serving as a classic case.

Core Data

  • FY2021 Revenue：60.12 billion RMB (publicly available data caliber, independent review unverified)
  • FY2026 External Commercialization Revenue Growth：45% (22-quarter high) (publicly available data caliber, independent review unverified)
  • Hong Kong Customer Count：Over 50,000 (publicly available data caliber, independent review unverified)
  • Data Center Delivery Period：100 days (publicly available data caliber, independent review unverified)
  • Data Center Cost Reduction：Over 10% (publicly available data caliber, independent review unverified)
  • 2026 China Smart Cloud Market Share：35.8%-38% (industry leader) (publicly available data caliber, independent review unverified)
  • Single-Node Queries Per Second：510,000 (publicly available data caliber, independent review unverified)

Competitors / Peers

Alibaba Cloud's main competitors in the Chinese market are Tencent Cloud, Huawei Cloud, and telecom operator-affiliated clouds. Tencent Cloud binds top internet clients through gaming, video, and the WeChat ecosystem, with strengths in PaaS and audio-video capabilities. Huawei Cloud has scaled rapidly since 2017 leveraging government-enterprise channels and domestic chip synergies, forming direct competition in government and manufacturing industries. Alibaba Cloud's advantages lie in its e-commerce ecosystem spillover and self-developed full stack, but it faces pressure from Huawei Cloud in government-enterprise payment collections and private cloud customization levels. The explosion of training and inference computing demand brought by generative AI in 2026 has extended the price war between Alibaba Cloud and operator clouds from general-purpose computing to GPU clusters and model calls, loosening the competitive landscape once again.

Sources

Related

Same industry / scale
Conglomerate / Trading HouseChina

Alibaba: From Chinese E-commerce Giant to Global Cloud Computing and Digital Economy Ecosystem Leader

Conglomerate / Trading HouseMulti-region

Samsung SDS: Transitioning from an Internal Group IT Service Provider to a Global Logistics and Cloud Player

Conglomerate / Trading HouseGlobal

CJ Group: From Samsung's spin-off sugar refinery to a global cultural empire built on K-food and entertainment

Conglomerate / Trading HouseGlobal

Reliance Industries: From Polyester Mill to Indian Conglomerate in Telecom, Retail, and Energy

⚠️ This entry draws on self-reported accounts and public reporting, not independently verified, and is not investment, startup, or legal advice; revenue figures are case-specific and not representative of typical results.