Price Increases Aren't Optional—They're Essential: JD Cloud's Awkward Position

In the business world, price has never been just a number; it is a thermometer of supply and demand, and a gauge of industry chain status.

When Alibaba Cloud and Baidu Smart Cloud begin raising prices, while JD Cloud chooses to “cut prices against the trend,” this is not simply a matter of differing pricing strategies but a naked display of power dynamics within the cloud computing industry—those able to raise prices are infrastructure providers, while those unable to do so are only edge services.

When costs generally rise, daring to increase prices means having the ability to pass on costs, while forced price cuts often reveal a passive position in market competition.

This seemingly ordinary price adjustment actually uncovers the real landscape of China’s cloud computing market after entering the deep water zone: the head effect is intensifying, the Matthew effect is becoming more apparent, and companies that cannot control pricing power are gradually being pushed to the lower end of the value chain.

The truth behind not raising prices:

It’s not about giving discounts, but lacking bargaining power

On the surface, JD Cloud’s “not raising prices or even lowering them” seems to be a gesture of industry concession, sharing the pressure of rising computing costs with customers.

This narrative may win some public relations favor temporarily, but when analyzed deeply within the industry competition landscape, this move appears highly passive. The essence of cloud computing is a large-scale infrastructure business with a high proportion of fixed costs.

Once demand explodes and supply tightens, price increases should be a natural outcome. Especially amid exploding AI computing needs, costs for GPUs, storage, and network bandwidth continue to rise, and expenses for power and cooling systems are also increasing. Raising prices indicates a tight supply-demand relationship and resource scarcity—typical features of a seller’s market.

In such an environment, choosing not to raise prices usually only has two possibilities: either possessing a strong cost advantage—like early Amazon Web Services leveraging extreme economies of scale to lower marginal costs—or lacking bargaining power, only able to compete on price to retain customers and prevent churn. From the current landscape, JD Cloud clearly leans toward the latter.

Tracing JD Cloud’s roots, it evolved from JD Group’s retail and logistics system. This background gives it unique advantages in supply chain scenarios but also limits its expansion in the general public cloud market. Compared to Alibaba Cloud and Baidu Smart Cloud, JD Cloud’s external customer structure is relatively simple, with weaker ecosystem binding capabilities, and its technological barriers in general-purpose computing have yet to form an absolute moat. This means it is not a “price setter” but a “price follower.” When leading players build defenses through technological premiums and brand effects, followers can only wield price as a double-edged sword.

In essence, not raising prices is a passive defense, not an active concession. When upstream hardware costs rise but cannot be passed downstream, the difference can only be absorbed by the cloud vendor’s own profit margins. Over time, this leads to limited R&D investment, weakened technological competitiveness, and a vicious cycle of “thin profits - less investment - weak products - only price cuts.”

In the winner-takes-all industry of cloud computing, lacking pricing power not only compresses profit margins but also risks marginalization in long-term competition.

Why do Alibaba Cloud and Baidu Smart Cloud dare to raise prices?

Reconfiguration of supply and demand in the AI era

Compared to JD Cloud, the price hikes by Alibaba Cloud and Baidu Smart Cloud are more like inevitable results of industry maturity and the release of benefits from technological cycle shifts. The core reason is simple—AI has changed the supply and demand structure of cloud computing.

Over the past decade, the core of cloud computing has been “virtualized resources,” mainly selling standard units of compute, storage, and networking. It was an era of severe homogenization and frequent price wars, with long-term price declines, where vendors competed mainly on squeezing margins.

But in the AI era, the essence has shifted to “computing resource allocation.” Graphics processors (GPUs) have become the core production material, with limited supply and exponentially growing demand driven by large model training and inference. This gives cloud vendors their first taste of pricing power akin to the “energy industry.”

In this context, price increases are not just cost pass-throughs but a business model upgrade—cloud is no longer a low-margin infrastructure but a high-value AI productivity platform. Alibaba Cloud relies on the Tongyi Qianwen large model, Baidu Smart Cloud on Wenxin Yiyan; they no longer just provide servers but offer full-stack services for model training, fine-tuning, and inference. These services involve extensive software optimization, algorithm tuning, and technical support, with value far beyond hardware alone. Customers are buying not just cold machines but intelligent problem-solving capabilities.

More critically, leading cloud providers have completed ecosystem locking. Once a company builds data, models, and business processes on a cloud platform, migrating is costly. The law of data gravity applies: data stays where the compute and applications are.

When a company deploys core business systems, data lakes, and AI models on Alibaba Cloud or Baidu Smart Cloud, switching providers entails huge reconstruction risks and time costs. This grants cloud vendors strong customer stickiness.

Therefore, the confidence to raise prices comes from two levels: resource scarcity and ecosystem binding. In the face of GPU supply constraints, whoever can provide stable, high-performance compute power holds the bargaining chip. Leading vendors, through years of accumulation, have built a complete chain from chip adaptation to framework optimization and model services—this deep integration is hard for latecomers to replicate quickly. Price increases are, in fact, market recognition of this scarcity and integration capability.

The layered era of cloud computing:

Capital markets only reward “companies that can raise prices”

This wave of differentiation is pushing the cloud market toward a clearer structure: top-tier cloud providers become “infrastructure providers” in the AI era; edge cloud vendors gradually become “price competitors.” This stratification is reflected not only in market share but also in valuation logic.

From an investment perspective, this stratification is significant. Past doubts about Chinese concept stocks often centered on their lack of sustainable profitability and pricing power. Investors grew tired of stories focused solely on scale without profit, shifting attention to free cash flow and return on net assets. Cloud computing, especially AI-driven cloud services, is one of the few sectors capable of providing “long-term cash flow with high growth.”

This is why companies like Alibaba and Baidu are re-establishing valuation anchors through their cloud businesses. When cloud can generate independent cash flow and has pricing power, it ceases to be a cost center dragging down group profits and instead becomes a growth engine for valuation recovery. Wall Street and Hong Kong investors are now viewing Chinese top cloud firms through the lens of Amazon Web Services or Microsoft Azure, focusing on profit margin improvements and revenue quality.

Conversely, if a business like JD Cloud cannot establish pricing power, even with revenue growth, it’s hard to achieve valuation premiums. Without profit support, revenue growth is fragile—especially during macroeconomic fluctuations.

If a company can only grow by lowering prices, the quality of that growth is questionable. The harsh but clear logic is: in infrastructure industries, price is not a result but a form of power. Having pricing power means resilience to economic cycles, the ability to pass on risks, and the capacity to set industry standards.

Moreover, this stratification influences talent and technology flows. High profits mean high R&D investment, leading to technological leadership, which further consolidates pricing power. Leading firms are forming a positive feedback loop, while those lacking pricing power risk resource depletion.

In the coming years, China’s cloud computing market may undergo a reshuffle: small and medium players unable to find differentiation in the AI era or establish pricing power will either be acquired or exit the market.

Conclusion: Pricing power,

The ultimate dividing line in the cloud computing industry

JD Cloud’s refusal to raise prices is not a competitive advantage but a signal—it has not yet entered the core layer of AI infrastructure. It remains in the stage of competing for existing market share through pricing, without establishing an irreplaceable value anchor in the growth market.

In this era, the market will ultimately reward only two types of companies: those that control compute power, and those that control pricing power. Controlling compute means owning scarce production assets like high-performance GPU clusters; controlling pricing power means having unique ecosystem barriers, such as irreplaceable model services or extremely high switching costs. Without both, any “discounting” will not lead to a true valuation reset.

This price divergence is, in fact, part of the industry’s coming-of-age—marking the end of the era of reckless growth and the beginning of the era of value return. For investors, identifying who holds pricing power is more critical than simply comparing revenue growth, because over long cycles, only companies with pricing power can weather fluctuations and turn technological dividends into lasting business value.

The second half of cloud computing is no longer about who is cheaper, but about who is more indispensable.