While working on some cloud topics I dig deeper in some core elements of cloud computing. One core element is the reflection on economical questions.
When thinking about it, it became clear that there are mainly to perspectives, a macro and a micro perspective. Both needs to be taken in account when trying to understand the economical side of the cloud.
The Macro Perspective
In its essence you can perceive the whole cloud computing area as a mainly economical topic. You may reject this statement and come up with quite a lot technical topics, that make cloud computing great. That can be things like high availability, geo redundance, elastic and flexible scaling of resources, accessibility from anywhere, availability of virtually any technology on short notice.
All those come true, definitely. But all those can be accomplished with your infrastructure as well. Surely, there is huge innovation provided by the big players in the cloud computing market, especially the IaaS (Infrastructure as a Service). But most of this is related to the massive scaling needed to provide services to thousands of customers.
Don’t get me wrong here, cloud computing will be a major corner pillar for information technology. It has started and it is growing.
But while looking from a technological perspective you can achieve all that benefits within in your own IT facilities, you will run into major issues. Most of that issues will be technology related, but can be overcome with enough budget and the right people.
Now that is why the cloud is in its essence an economical questions. You may not have enough budget and the right people. Economical reasons will prevent you from building the needed technology, economical reasons will also prevent you from finding the right people, build the right teams and give them the freedom to stack up all the needed knowhow. It is simply not economical to do this all alone.
Secondly to be able to scale as needed you need to have enough resources, Again a lot of costs to have on the list.
The big players spread that costs over a lot of customers. Massive scaling.
The Micro Perspective
When evaluating the deployment of workloads in the cloud, often the question is whether a company should stay on-premises or should go into the cloud.
Now while this simplified macro view is not really difficult to understand, people often narrow down their decision whether to go for the cloud or not, mainly on economical reasons. Fine so far, but what we often see is that they miss a lot of important factors.
During the process this is the used perspective, so people tend to compare costs for running workloads in the cloud or on their own servers. Very often the result indicates the cloud might be much more expensive.
How can that be?
When a cloud provider sets the price tag, this includes everything needed to run for example a server instance. This includes costs for IT, facilities, management and perimeter costs.
- IT costs are mostly related to hardware, software, installation, power consumption, networking and peripherals.
- Facility cost include housing, racks, ventilation and air conditioning, emergency power supply, fire protection, building, property, etc. And all that needs to be maintained and from time to time it needs to get cleaned.
- Management involves all aspects of running your systems, maintain and operate it.
- And the there are still more costs, what I call perimeter costs. Personell, that is not directly involved in the service, but is needed, like security personell, you need insurances, parking space for staff etc.
This are all factors, that are included in the price of an instance by the cloud provider of your choice.
Now for comparison people calculate prices for running instances on their own. Usually here costs for IT is included, but usually only for hardware and software. If you need some new periphereal hardware like networking equipment, it is included as well. Power consumption is also part of the equation. But usually thats it. All the other factors are seldom taken into consideration. Just because it is already there. You already have a building, you have rooms, you have ventilation, you have a security guy, the building is already getting cleaned by somebody, and your networking and perimeter supplies are also already there.
From this perspective the cloud often seems to be more expensive, but only because you exclude a lot of factors from the equation.
So in the essence, most IT managers have the micro perspective on their radar, but does not reflect all the factors.
Another reason is that often only a small piece is evaluated, but the real value comes clear, when you look at the whole picture. You will not reach a critical mass, when you evaluate each application on its own.