"The pain of a curve switch is, in essence, the 'price of growth,' not a 'signal of failure.'"
Recently, Huawei Cloud's organizational restructuring has drawn widespread industry attention — it disbanded the Cloud EI (Enterprise Intelligence) product line and consolidated its business around the "3+2+1" system (three foundational capabilities — general computing, AI computing, and storage; two core platforms — AI PaaS and databases; and one safeguard business — security). Unofficial information indicates this adjustment involved personnel changes for nearly a thousand people. Many interpret it as "business contraction," but seeing the essence through the phenomenon, this is not a "decline signal" for Huawei Cloud; rather, it is the inevitable pain of a company switching from its "second cloud-computing curve" to its "third AI curve." This pain is no special case for tech giants alone — it is a strategic proposition every company trying to cross a technology cycle must face.
The internet broke the spatiotemporal limits on information dissemination and interaction; cloud computing changed how computing resources are delivered and used; and AI strives to endow machines with intelligence for more efficient decision-making and services.
01 The First Principle: Deconstructing Business Essence with the Growth-Curve Model
Before discussing Huawei Cloud's adjustment, we must first clarify the underlying logic of the tech industry: the iteration of technology waves is essentially the replacement of growth curves — the internet era centered on "information-connection efficiency," the cloud era focused on "computing-delivery efficiency," and the AI era has shifted to "intelligent-generation efficiency." For companies, the core contradiction of a curve switch is not "whether to transform" but "how to efficiently migrate the old curve's resources (technology, teams, customers) to the new curve" — which is precisely the first principle of technology-driven companies:
The essence of business growth is the efficiency of resource reallocation; the pain of a curve switch is, in essence, the inevitable cost of mismatching old resources with new demands.
In the internet era, the key to corporate competition was achieving large-scale traffic through search engines and social platforms; entering the cloud era, the focus shifted to turning computing resources into "on-demand public services" — Huawei Cloud's early deployments of elastic computing and government-enterprise dedicated clouds were precisely the implementation of "computing-delivery efficiency." Today, under the AI wave, the dimension of competition has upgraded to "the ability to use intelligence to solve real problems," such as providing AI quality-inspection solutions for manufacturing or developing imaging-analysis systems for healthcare — a business logic completely different from traditional cloud computing.
The contradiction of resource mismatch thus stands out: Huawei Cloud's former Cloud EI department focused on "tooling enterprise intelligence," such as data visualization and basic algorithm-model development, whose capability system differs significantly from the AI-era demands for "generative native architecture" and "industry deep-solution"; old-business investment struggles to yield new returns, and old-team capabilities struggle to fit new tasks, making organizational restructuring and staff optimization necessary moves to balance resource allocation and reduce mismatch costs.
This kind of adjustment is not unique: when Amazon transformed from e-commerce to AWS cloud computing, it cut many traditional e-commerce tech roles to focus on cloud-service R&D; when Google built up DeepMind to advance AI, it also restructured the resource allocation of its advertising business — every company crossing a cycle must pay the tuition of "resource mismatch." Pain is the standard equipment of a curve switch, not the exception.
02 Huawei Cloud's Pain: The Common Challenge of Every Company's Strategic Transformation
Huawei Cloud's adjustment is not "blind transformation" but is driven by three real constraints — profit targets, industry competition, and capability boundaries. These three constraints are also checkpoints every company must pass when transforming toward AI; it is only that Huawei Cloud, as an industry front-runner, faced and addressed them earlier.
1. Profit-Target Constraint: From "Curve Extension" to "Curve Connection," Divesting Non-Strategic Assets
In 2024 Huawei Cloud was still in the red, and "achieving profitability" was set as a clear core goal for 2025. This goal forced the business to shift from "scale coverage of a single curve" to "cross-curve value focus": previously Huawei Cloud's business spanned Cloud EI tools, edge computing, large AI models, and more, with some businesses (such as basic algorithm tools) facing dual pressure from "customer self-substitution" and "low-price competition," and continuously declining input-output ratios; this disbanding of the Cloud EI department and focus on the "3+2+1" system is, in essence, divesting "non-strategic assets" and concentrating manpower, capital, and other resources on high-value areas such as AI computing and AI PaaS — by narrowing the business boundary, it improves the precision of resource investment and shortens the profit cycle.
The essence of this series of moves is the company's "curve relay": like a relay race, the resources of the second leg (cloud computing) must be efficiently passed to the third leg (AI); if the second leg's "redundant resources" (Cloud EI tools) slow the pace, they must be decisively dropped to secure the overall rhythm. For Huawei Cloud, what it "drops" is the low-value tooling business, and what it "focuses on" is the AI core capability that can support long-term profitability.
2. Industry-Competition Constraint: From "Waiting on Technology" to "Curve Positioning," Seizing the Window
The current cloud-computing market has entered a "stock competition" stage (the maturity phase of the green curve): Alibaba Cloud and Tencent Cloud consolidate share through price wars in the government-enterprise market, while AWS and Azure increase investment in AI computing — the growth space of traditional cloud business keeps narrowing; at the same time, the window of AI technology is closing fast — the compute density of AI clusters and the industry-adaptation ability of AI models have become the core competitiveness of cloud providers. If Huawei Cloud clings to traditional cloud business or waits for AI technology to be "fully mature" before transforming, it will miss the market's first-mover advantage.
The balance between business realism and technological idealism is highlighted here: technological evolution can wait for an "ideal state" (such as the "super task-execution unit" predicted to appear in 2026–2027), but corporate competition cannot wait. Huawei Cloud's adjustment is, in essence, "trading short-term organizational cost for long-term curve positioning" — by integrating resources ahead of time to focus on AI, it builds advantages in areas such as AI-computing infrastructure and industry AI solutions, avoiding the dual predicament of "stagnant traditional-business growth and missed first-mover advantage in new business."
3. Capability-Boundary Constraint: From "Full-Spectrum Layout" to "Curve Trade-offs," Anchoring Core Advantages
Huawei Cloud's core advantage lies in the synergy of "hardware + government-enterprise resources" (the core assets accumulated on the second curve): it can reuse the computing advantage of Huawei's Ascend chips and leverage its government-enterprise customer base to advance industry AI adoption; but its capability boundary is equally clear — in areas such as C-end AI applications and general large models, it faces fierce competition from ByteDance, Baidu, and others, and can hardly reuse its own core resources. This focus on the "3+2+1" system is, in reality, a clarification of its capability boundary: not entering C-end AI consumer scenarios, concentrating on B-end industry solutions, and reducing transformation risk through "playing to strengths and avoiding weaknesses" (the synergy zone of the blue curve).
The core reason many companies fail in transformation is precisely "blind expansion that breaches capability boundaries" — wanting both to build large models, develop AI tools, and break into consumer applications, leading to scattered resources and dissipated energy. Huawei Cloud's choice is to "first hold the capability boundary, then expand gradually"; though it bears the pain of "abandoning temptations," it gains the certainty of a "focused domain."
03 Crossing the Valley of Death: The Common Challenge of Every Company's Strategic Transformation
The Huawei Cloud case offers a reusable strategic framework for all companies facing a curve switch. Whether manufacturing, retail, or technology, every industry will encounter the transformation challenge of "from the second curve to the third curve" in the next five years; the key is to make three major strategic choices that turn pain into growth momentum.
1. Resource-Anchor Choice: Use the Old Curve's Core Capability to Build the New Curve's "Migration Ladder"
Transformation is not "starting from scratch" but "new reuse of old resources." Huawei Cloud's resource anchor is the computing infrastructure (data centers, AI clusters) and government-enterprise customer resources accumulated in the cloud era — when advancing AI business, there is no need to rebuild the computing base; existing data centers can be reused directly; there is no need to re-acquire customers; AI quality inspection, smart government affairs, and other solutions can be advanced based on existing government-enterprise cooperation. This "reuse of old resources" greatly reduces transformation cost and risk.
For traditional enterprises, the choice of resource anchor is equally critical: a retailer transforming with AI should anchor on "member data" and "store scenarios," not blindly build large models — by optimizing AI recommendation algorithms with member data and developing AI foot-traffic analysis systems with store cameras, it both reuses existing resources and rapidly generates business value; a manufacturer transforming with AI should anchor on "production data" and "equipment assets," focusing on scenarios like AI quality inspection and intelligent scheduling, rather than pursuing "general AI capability." The core of a curve switch is to "cut old business, not discard old resources," letting the old curve's accumulation support the new curve.
2. Survival-Threshold Choice: Balance Short-Term Cost and Long-Term Profit, Avoid "Dying from Transformation"
Huawei Cloud's adjustment is a rational decision based on "cost calculation" — the short-term cost of layoffs and business integration (severance, organizational friction) is lower than the long-term drain of "continuing to maintain low-value business," and the company's financial strength is sufficient to sustain until the AI business turns profitable. But for small and medium enterprises, they must avoid "copying the giants' moves" and instead choose a pain threshold of "small-step trial and error": first focus on one or two high-frequency scenarios (such as intelligent customer service or AI expense reimbursement) for pilots, validate business value and accumulate transformation experience through the pilots, then gradually expand the scope.
The core of threshold choice is "cash-flow balance": the correspondence between "transformation investment − short-term output" must be clear — if 1 million yuan is invested in AI transformation and the pilot scenario achieves 500,000 yuan in revenue with a clear growth trend, proceed; if the pilot yields only 100,000 yuan in revenue or a payback period exceeding two years, re-evaluate the direction. Pain is necessary, but it must not breach the "company's survival bottom line"; short-term cost must match long-term profit expectations.
3. Battlefield-Focus Choice: Lock in "Winnable Battles Within Capability," Reject "Hot-Track Temptation"
Huawei Cloud abandoned the C-end AI market because it recognized that the "government-enterprise B-end" is the "must-win battlefield" within its capability; if it forcibly entered the C-end, it would not only struggle to reuse its resources but also face internet giants head-on, possibly ending up "losing more than it gains." When companies transform with AI, they must likewise avoid "chasing hotspots and grabbing tracks" and instead focus on "niche scenarios within capability": a medical-device company should focus on "device-end AI diagnosis" rather than general large models — it can develop AI-assisted diagnosis on existing medical devices, reusing medical customer resources and device data to form a differentiated advantage; a logistics company should focus on "AI intelligent dispatching" rather than consumer AI applications — by optimizing transport routes and intelligently matching orders, it quickly improves logistics efficiency and profitability.
The pain of a curve switch is, in essence, the "pain of growth," not a "signal of failure." Huawei Cloud's current adjustment is like a tree pruning in winter — shedding redundant branches so that sturdier new shoots sprout in spring; short-term organizational optimization is to seize a more favorable competitive position in the AI era.
Huawei Cloud's present may well be the tomorrow of countless companies. In the iteration of technology waves, pain is not an endpoint but the starting point of the next growth curve — only by embracing pain and mastering mismatch can a company's growth curve keep extending upward through the tides of the era.
When the old curve's resources become a drag on the new curve, do you dare to "let go"?
When the industry's window is closing fast, can you "seize the position"?
When capability boundaries meet temptation, will you "stay in your lane"?
The answers to these questions determine whether a company can metamorphose into a true cycle-crosser in the valley of death of curve switching.
