The Economist this week (September 3, 2022) carried an article titled, ‘Cloud Computing – Margin Brawl’. It is an excellent article that provides a neat structure to explore how cloud computing is evolving. In fact it’s sections provide a nice way to ask critical questions and explore the options.
The first part of the paper shows how the large Cloud Service Providers or CSPs (really, the cloud infrastructure providers that offer compute and storage) such as Google, Amazon, and Microsoft have recently signed up clients for their financial-services cloud. The main point of the paper is stated: as competitors increasingly offer similar services, won’t such behavior lead to falling margins due to increased completion? This is a key question for all vendors who offer services in the cloud. Not least because even Gartner suggests the cloud is part of our collective future.
Where the Cloud Came From
The article than steps back in time and shows that in fact the main CSPs have historically targeted slightly different markets. Only recently have they started to compete head to head more directly. If you throw in some other not-so-cloudy ‘mega vendors’ like Oracle and SAP, and others like SalesForce, you get a feeling that all these vendors are destined to bash each other up: to partner, acquire, or change direction. The paper asks the question – when will margins fall?
Most recently reported profits have been healthy. As such there seems little pressure on margins. But the paper then suggests that the smart boffins at the larger CSPs see the future and declining margins, so they have increasingly been trying to move up the stack. This means they are offering more and more software, thus partnering and/or competing more with the likes of SAP, Oracle and SalesForce, and other Independent Software Providers, or ISVs. When partnered, the marketing word ‘ecosystem’ is mentioned and so the paper does so in the right place. The longer term market result is fewer overall options (ecosystems) but more pre-packaged and pre-integrated software and business services.
Confusingly the paper then introduces the term ‘hyperscalers’ and I believe it is used to describe the largest CSPs. But the reader is left to figure this out for themselves.
From Compute and Storage to Software and the Stack
The stack is then explored again, from the chip and hardware at the bottom, to the software running at the top. The stack is discussed as a source of scalable differentiation and margin. Software is eminently more profitable than hardware. Software can be copied and reused; hardware cannot. Vendor lock-in is discussed and several recent trends suggests that vendor lock-in as a risk is falling. Egress fees, charged by CSPs (or hyperescalers?) to move a firms’ data to a different cloud (CSP, or hyperescaler), are falling.
What the paper then does is explore or relate the stack and each layers contribution to innovation and productivity. Hardware, as in storage and computer, was the foundation of cloud computing. It’s what the market calls Infrastructure as a Service (IaaS). Effectively this is already a near commodity and the hardware used to power it has been subject to Moore’s Law. This helps explain the critical role the High Tech industry has in contributing to overall economic productivity. As firms in other industries use IaaS, they too get a fillip in productivity growth. But IaaS is pretty much well established.
Software is quite different to hardware. Software, data and IP, are what is used to describe what a firm does, not what computing hardware does. These include business processes, analytics and decision making. IaaS vendors thus have added Software as a Service (SaaS). Software can be copied and adapted, and that’s called Platform as a Service). PaaS is what provides unique software-based business processes and decision support tools.
Who Dares, Wins
SaaS has the higher margins over PaaS, and PaaS has higher margins than IaaS. Hence the race to the stop. But the interplay of the competitive forces and market dynamics are not really explored in the paper. The paper however does make a very important connection at the end. It seems the race to the top might not end well for CSPs. Insightfully the paper argues that regulators will not take too kindly to these CSPs and their ecosystem of ISVs dominating the whole stack. This is the area I have been looking at and what you see playing out with the Apple Store and Epic could well play out in cloud computing. See Keep an Eye on the Apple v Epic Battle.
The EUs Digital Market and Digital Services Acts, which will start to come into force next year, may yet signal that The EU will cast their gaze in this direction. I think it will. As such the race to the top could well be the straw that breaks the clouds up. Time will tell.