The cloud giants stumble

Recent financial results from the major cloud providers reveal a concerning trend in cloud adoption and growth. AWS reported just 13% growth in the final quarter of 2023, Microsoft Azure saw 19% growth, and Google Cloud achieved 26% growth—all below market expectations. The slowdown is particularly noteworthy as these three cloud titans collectively control 67% of the $74 billion global cloud infrastructure market.

The third-quarter 2024 results show continued struggles, with AWS posting 19% growth, Microsoft Azure at 20%, and Google Cloud at 35%. These numbers have left investors questioning the providers’ massive AI investments. Infrastructure bottlenecks force these companies to invest billions in data centers with little short-term payoff.

Why is growth slowing?

The lackluster performance can be attributed to several factors. Enterprise customers are increasingly repatriating applications and data back to on-premises environments, driven by the mounting costs of public cloud services. The notorious egress fees and other hidden costs have created significant friction, leading organizations to reevaluate their cloud strategies.

In my opinion, enterprises are becoming more sophisticated in their approach to cloud computing. Many organizations have discovered that the promised cost savings of the public cloud often fail to materialize, especially for steady-state workloads. The “lift and shift” approach, once touted as a quick path to cloud adoption, has proven to be more complex and expensive than initially projected.

The emergence of specialized AI providers and microclouds is fragmenting the market further. Companies like CoreWeave have demonstrated that focused, specialized infrastructure can deliver better performance and value for AI workloads than general-purpose cloud platforms. This trend is particularly evident in machine learning and deep learning, where specialized hardware configurations and optimized environments are crucial.

Another significant factor is the growing awareness of data sovereignty and control. Organizations are increasingly concerned about data governance, privacy, and compliance requirements, leading them to hybrid or private cloud solutions. The major cloud providers’ one-size-fits-all approach is proving less attractive as enterprises seek more nuanced solutions that align with their regulatory and operational requirements.

Providers don’t see their problems with AI

I wrote about this recently, so I won’t go into depth here. The AI ecosystem development by significant cloud providers has also fallen short of expectations. Providers have substantially invested in AI infrastructure and services, but companies are finding that the costs of building and maintaining AI models on these platforms are difficult to predict and control.

Moreover, the rise of edge computing and Internet of Things applications has highlighted the limitations of centralized cloud computing models. Processing data closer to its source reduces latency and can significantly cut costs associated with data transfer and storage. This realization has increased investment in edge computing infrastructure and private cloud solutions.

These results serve as a wake-up call for the cloud industry. The days of unchallenged public cloud dominance may wane as enterprises become more knowledgeable about their infrastructure choices. Rather than driving growth, AI investments are currently creating capital-intensive challenges for the major providers.

A changing cloud market

Looking ahead, the cloud computing landscape is likely to become more diverse and specialized. Although the major cloud providers will continue to play a significant role, their dominance may diminish. The success of specialized providers and the growing preference for private and hybrid solutions suggest that the future of cloud computing will be more distributed and heterogeneous than the current model. But that’s not new; I’ve been saying that for years.  

The challenge for Amazon, Microsoft, and Google will be to adapt their strategies to this evolving landscape. They’ll need to address concerns about costs, provide more flexible deployment options, and develop compelling AI solutions that deliver clear value to enterprises. Without these changes, they may continue to see their growth rates decline as organizations increasingly turn to alternative solutions that better meet their specific needs. This does not mean failure for Big Cloud, but they will take a few years to figure out what’s important to their market. They are a bit off-target now.

The rise of specialized providers and the growing acceptance of private cloud solutions means enterprises can be more selective, choosing fit-for-purpose options rather than forcing all workloads into a one-size-fits-all public cloud model that may not be cost-effective. This is particularly relevant for AI initiatives, where specialized infrastructure providers often deliver better value.

This freedom of choice comes with increased responsibility. Enterprises must develop more substantial in-house expertise to effectively evaluate and manage multiple infrastructure options. They’ll need to sharpen their skills in cost management, workload optimization, and multi-cloud orchestration.

The key takeaway is clear: Enterprises are entering an era where they can build infrastructure strategies based on their specific needs rather than vendor limitations. This shift promises better economics and more control but requires enterprises to be more sophisticated in their decision-making.