In my recent post What Happens When a Fixed Cost Turns Variable I discussed what happens when a company benefiting from economies of scale over a fixed cost experiences that cost turning variable. To summarise: bad things happen.
But the reality is that most companies do not begin with economies of scale over any, let alone several, of their costs. In these situations it makes sense to outsource one’s costs as much as possible.
Being able to outsource one’s major IT-related costs is at the core of the cloud’s customer value proposition to customers.
Companies shifting to the cloud are turning the fixed, upfront capital cost of servers and other IT-related expenses into a variable operating cost. Concurrently, the reverse is happening for the cloud providers themselves, who are developing ginormous economies of scale & scope in providing cloud services to their customers.
These ‘Big Three’ cloud providers (Amazon Web Services, Microsoft Azure, and Google Cloud Platform) are commoditising servers and compute, and selling the service of: performance; scalability; reliability, and security.
Structure of this post:
- Back-to-Basics - What is the cloud?
- Who are the major cloud providers?
- What is the value proposition for customers in moving to the cloud?
- Why is there a secular shift towards the cloud?
- Where are we in the shift to the cloud?
- Why are the Big getting Bigger?
- Areas for further thought & research
1: Back-to-Basics - What is the Cloud?
Here is how Amazon Web Services defines cloud computing:
Cloud computing provides a simple way to access servers, storage, databases and a broad set of application services over the Internet. Cloud services platforms own and maintain the network-connected hardware required for these application services, while you provision and use what you need via a web application.
With cloud computing, you don’t need to make large upfront investments in hardware and spend a lot of time on the heavy lifting of managing that hardware. Instead, you can provision exactly the right type and size of computing resources you need to power your newest bright idea or operate your IT department. You can access as many resources as you need, almost instantly, and only pay for what you use.
Cloud computing is the on-demand delivery of compute power, database storage, applications, and other IT resources through a cloud services platform via the Internet with pay-as-you-go pricing. Whether you are running applications that share photos to millions of mobile users or you’re supporting the critical operations of your business, a cloud services platform provides rapid access to flexible and low-cost IT resources.
Thanks to the cloud, many companies and households today get their IT needs in the same way that all companies/households get their water and electricity:
- Companies and households do not own their own waterworks or power plants but instead access them on-demand by turning a tap or flicking a switch
- Fewer & fewer companies/households buy, house, or maintain their computing environment (i.e. data centers and servers), but instead access these resources over the internet with pay-as-you-go pricing just by clicking a button
The essence of the cloud is in providing compute processing, storage, applications, systems management, and security all over the internet as an on-demand service - much in the same way as we access water & power on-demand.
You can think of it as ‘utility computing’.
2: Who are the major cloud providers?
The major providers of cloud computing services are, in order of market share:
- Amazon Web Services (AWS)
- Microsoft Azure (Azure)
- Google Cloud Platform (GCP)
Gartner’s magic quadrant places for providers of Cloud Infrastructure as a Service is as follows:
For context Amazon generated $35bn of revenue from AWS in 2019, up 37% from $26bn the prior year. Gartner estimated that AWS had ~50% market share in Infrastructure-as-a-Service compared to Azure’s ~15% in 2018. Azure and GCP gained market share in 2019, but AWS remains the undisputed leader.
Many companies today have cloud offerings but really use these one or more of the Big Three behind the scenes. For example, Guidewire has a cloud offering for its Property & Casualty insurance customers, but Guidewire’s cloud offering runs on top of Amazon Web Services:
Source: Guidewire company presentation to investors.
When you hear of ‘cloud-native’ companies, it doesn’t necessarily mean that they own and operate the servers themselves. They are typically running a specialized service on top of the core cloud infrastructure provided by one of the Big Three, much like Guidewire. ‘Cloud-native’ is just a buzz-phrase for being created to run on the cloud as opposed to being initially created to run on-premise.
Note that there are some direct competitors to the Big Three’s services, but these tend to be fringe competitors who are competing over specific use cases.
3: What is the value proposition for customers moving to the cloud? i.e. why doesn’t everyone just do this themselves on-premise?
There are several reasons why this idea of ‘utility computing’ is a compelling proposition for customers. Customers:
- Don’t have to own as many assets and can convert lumpy, upfront, fixed capital expenses into variable cost of revenue
- Gain access to unlimited computing resources on an as-needed basis i.e. greater ‘scalability’
- No longer have to endure the headaches of technology cycles, upgrades, maintenance, integration, and management
- Outsource other facets of IT that they are not domain experts in - including performance, security, and reliability
Let’s briefly go through each of these:
What’s the value proposition of having fewer assets and converting fixed capital expenses into variable operating costs?
Cloud computing lets you focus on your business rather than on the heavy lifting of racking, stacking, and powering servers.
Before the advent of cloud computing you had to invest heavily in servers and even whole data centers before you knew how you were going to use them. If one of these servers stopped working your website/business could come crashing down until you fixed whatever was going wrong with your server (or until you hired someone to come fix it for you).
Today, you can sign up with AWS/Azure/GCP in minutes, pay them only when you consume computing resources, and pay them only for how much computing resources you consume.
This ease of sign up allows you to deploy your application in multiple regions around the world with just a few clicks. It also allows developers to access resources in minutes that might have previously taken months. This is a massive cost saving that is simultaneously difficult to quantify and difficult to overstate.
Another benefit of using AWS/Azure/GCP’s state-of-the-art servers & infrastructure is that they provide lower latency and a better experience for you at a lower cost than you likely could achieve with your own servers.
Finally, as we increasingly use our smartphones/smartwatches etc, it is a lot to ask that these devices store all our data while also functioning at high speed. It makes a lot more sense for your smartphone to access data over the internet (where the data is actually stored in some server farm in Virginia/elsewhere).
What’s the value proposition of access to unlimited computing resources on an as-needed basis?
It’s easy to grasp conceptually how beneficial it is to be able to access unlimited computing resources on-demand, and to only pay for what you use.
From AWS’s website:
Eliminate guessing on your infrastructure capacity needs. When you make a capacity decision prior to deploying an application, you often end up either sitting on expensive idle resources or dealing with limited capacity. With cloud computing, these problems go away. You can access as much or as little capacity as you need, and scale up and down as required with only a few minutes’ notice.
The benefit of this has never been more apparent than during the Coronavirus.
Think about Zoom, and then think about Expedia. The demand for their services have gone in completely opposite directions since February:
In both Zoom and Expedia’s case, this is real money:
- If Zoom had used all their own servers and hadn’t provisioned for the sudden & huge increase in demand that the Great Lockdown brought for their services, they likely would have been overwhelmed with latency issues when dealing with their 354% yoy increase in customers. This would have resulted in real money lost since they would have missed out on the opportunity to add many of these new customers and they probably would have frustrated existing customers (many of whom would have gone elsewhere such as to Microsoft Teams).
- If Expedia had paid upfront for all their regular server requirements, and then saw demand for their services fall off a cliff, that unused capacity sitting idle would be costing them money at a time when revenues have fallen off a cliff. To use a very simplified example, if IT costs would normally be 10% of revenues but are fixed, and revenues fall by 80% for two months, then suddenly those IT costs are 50% of your revenues during those two months of pain. It’s only a slight exaggeration to say that outsourcing this facet of their business to AWS has helped Expedia stay alive.
What’s the value proposition of shedding technology cycles, upgrades, maintenance, integration, and management?
Upgrading from one software version to the next is a pain in the derriere.
Upgrades cost time - time in the transition phase to ensure backward-compatibility with all other applications, time to fix whatever inevitably goes wrong in the transition, and time retraining people over even trivial changes in the user interface.
It also costs money - especially if you have to pay to license each new version of the software and/or pay someone to help you integrate the new system.
It’s far easier to have all this done for you automatically over the web, including maintenance and management of the servers.
What’s the value proposition of outsourcing facets of your business that you’re not an expert in, like performance, reliability and security to AWS/Azure/GCP?
Most companies simply are not domain experts in the areas where the Big Three are.
For example, a regional coffee chain should hardly be expected to have expertise in data security any more than they should be expected to have expertise in reliably generating electricity.
By outsourcing IT to a centralized player such as AWS, the coffee chain benefits from AWS’s expertise in data security as well as lower costs and all the other benefits listed above.
4: Why is there a secular shift towards the cloud?
There is a secular shift to the cloud for all the reasons mentioned above. Hopefully that is simple enough!
5: Where are we in the shift to the cloud?
We’re still early in this shift, and COVID-19 has accelerated the transition as individuals and businesses recognize the advantage of being able to (a) work remotely and (b) have variable, rather than fixed, IT costs.
Source: Gartner, I.E. estimates. Note that Gartner estimates were pre-COVID…
Many of the businesses that go out of business during the Great Lockdown will do so due to bloated cost structures. Shifting to the cloud would help take some of the bloat out of most businesses.
Clearly not every business using the cloud is fat-free and will survive this, but there is likely some causality in some staying in business due to having lower/more variable overhead costs during this period of unprecedented revenue contraction for many businesses.
More to the point, virtually every single company that starts from here on out will use the cloud to get started. As creative destruction takes its natural (and currently accelerated) toll, this will result in a greater share of businesses using the cloud in the future.
The big picture is that this is likely a multi-decade shift to the cloud that we’re currently in the early innings of (as of June 2020…).
6: Why are the Big getting Bigger?
What leads to the Big getting Bigger is that there are positive, reinforcing feedback loops which make these companies better at what they do over time. Simply put, these businesses benefit from massive economies of scale:
Usage from hundreds of thousands of customers is aggregated in the cloud. This enables cloud infrastructure providers to achieve higher economies of scale which translates into lower pay as-you-go prices for their customers.
As the Big Three add more customers, data centers & engineers and encounter more unique customer cases this enables them to build new solutions & increase the breadth of their offerings, making them more attractive to existing and prospective customers.
While there are limits to the benefits of the feedback loop that Starbucks baristas experience in making coffee for every additional customer that they serve, the feedback loop of AWS adding more customers leads directly to increased economies of scale and better service provided to all customers.
The scale this is happening on is staggering, largely because these companies are competing on a global (ex-China) basis.
Water & electrical utilities are natural monopolies, but only within their respective geographic regions.
In contrast, the Big Three’s economies of scale apply across borders.
For this reason we have possibly never-before seen economies of scale at quite this, well, scale.
There are undoubtedly benefits to building out data centers locally in terms of reducing latency, which matters a lot in cloud gaming, for example. But for most cloud applications having zero noticeable latency is less critical than, say, knowing that you’re partnering with a best-in-class cloud provider in terms of security, reliability and scalability (particularly given that the Big Three continue to outspend everyone else on capex, other than perhaps Facebook).
And note that the Big Three are very focused on edge computing. For example.
Competing with the Big Three at this point is a loser’s game due to the massive & increasing economies of scale & brand advantages that the Big Three already have. These competitive advantages are high and rising, and appear to be insurmountable already.
The customer value proposition of using AWS/Azure/GCP as your cloud infrastructure provider is compelling.
Simply put, these [warning: buzzword bonanza alert] hyperscaled cloud infrastructure providers have developed such enormous (and rapidly growing) economies of scale to the point where it’s fiscally irresponsible to not use one of them.
AWS, Azure and GCP are commoditizing servers and compute, and selling the service of: performance; scalability; reliability, and security. The more they commoditize servers, the more they can charge for their service.
8: Areas for further research:
- What is the best way to quantify the customer ROI from having access to AWS/Azure/GCP’s cost advantage?
- While economies of scale are undoubtedly here to stay for the Big Three, will there always be switching costs or will this become a commodity service where the low cost player wins?
- Does ‘Nobody ever gets fired for hiring AWS/Azure’ apply and is this a manifestation of Brand as a competitive advantage for the hyper-scaled players?
- Is this Conservation of Attractive Profits avant-la-lettre, or is this shift a step-change and a win-win for everyone?
- Is the logical end-game that the Big Three become regulated like utilities?
- When does it not make sense to switch to the cloud?