The Sunken Cost Fallacy of On-Premise Software

You can see the sunken cost fallacy in all walks of life. An example for me would be when I had tickets to see my football team play but felt absolutely awful. I went along because I’d already bought the ticket (sunk the cost) but, had a thoroughly horrible time because I was so ill. A more serious example was seen with the development of the Concorde, in fact, ‘The Sunken Cost Fallacy’ is often referred to as the ‘Concorde Fallacy’ because the project to build the Concorde ran so far over its budget and deadline that they knew it would never turn a profit, but they continued on development regardless.

concorde fallacy

As far as technology is concerned the sunken cost fallacy often rears its head. I knew of a wholesaler that spent millions on new hand held scanners to run the picking and packing within their warehouse. The scanners soon became incompatible with the wholesalers stocking software so the company had a choice: Buy new future proof scanners or pay a heavy cost to have their software adapted to work with the old hand-held PDA’s. They chose the latter based on the fact that they had spent so much on the original hand-held devices. As you expected the wholesaler was ok again for a while until the stocking software changed once more and they had to adapt the hand-helds again.

I see a close analogy of the wholesalers mistake occur often within the office of finance where the decision to painstakingly continue maintaining and upgrading an organisations hardware infrastructure is made based on the sunken cost fallacy that is because you are already heavily invested it is better to not change to a future proof cloud infrastructure.

Economists will state that when sunken costs influence business decisions it’s due to emotional thinking not logical, so how can you tell whether your decision to stick with an on-premise applications infrastructure is a logical decision or an emotional one? I’ve listed a few of my thoughts below:

1) Emotional Reason:- You’re invested. Perhaps an obvious one but the sunken cost fallacy goes deeper than just the money you’ve spent on physical hardware. It takes into account staff and their training, software licenses you may have bought, as well as money spent on integrations between other systems. This is an illogical reason because staff can be re-trained and deployed to other more productive roles, SaaS software is often cheaper, and providers like Oracle who are the 2nd largest SaaS company now offer their existing customers a chance to transfer some of their existing application investment into SaaS licenses, and integrations are often also transferable if they are even required anymore.

2) Logical Reason:- A logical reason to keep with your on-premise investment is because you have only VERY recently spent money upgrading the infrastructure. However, this is a slippery slope, if your organisation is able to migrate to a cloud offering that covers your needs then you should begin to plan for the transition. Otherwise you will keep making the emotional decision to keep upgrading your infrastructure based on the more recent sunken cost.

3) Emotional Reason:- You feel your data is more secure in your own environment. This is one of my favourite reasons, of course it is all entirely relative, and, I am sure there are some on-premise structures that are as secure as the most secure cloud structures, but likewise a tier 4 data centre is going to be more secure than your average on-premise structure. You need to ensure that the cloud provider offers single tenancy hosting. The video below gives a great insight into the old cloud vs the new:

4) Logical Reason:- The cloud doesn’t offer the functionality my organisation needs. This is a logical reason for remaining on-premise. Whilst the cloud has been around a long time the applications that run on them are often newer and occasionally don’t have the functionality required for some specialist businesses. In this scenario I recommend evaluating whether the short fall in functionality required outweighs the benefits that a SaaS offering provides.

5) Emotional Reason:- You made the decision before to stick with on-premise structure and nothing has gone wrong. A common misconception with on-premise infrastructures is that the cost only comes when buying new hardware. When compared to a cloud infrastructure you have to take in the running costs of expert staff, backups, electricity, not to mention things that are rarely considered like the physical space within your organisation that is taken up by the hardware systems.

6) Emotional Reason:- You want to keep control of your systems. This is the reason that borders closest onto being a logical reason, but, that is only because the lack of control is an issue that exists because some providers have shortfalls in their offering. The best of cloud providers will allow you a very good level of control over your systems and the benefits of smoother upgrades and hassle free maintenance outweighs any minor loss in control by a good cloud provider.

In summary my view is that the majority of reasons for not moving to the cloud are emotionally based rather than using logic. The benefits of the cloud are well publicised, and link closely with the goals of a modern business to increase agility, cut costs and empower people.

If there was ever a shining endorsement for the future of cloud it was that the head of development for one of the Worlds top software companies recently said that “a decision to go on-premise is just a delay in the decision to move to the cloud”.

What do you think? Have I missed something important out for or against on-premise? Please add your thoughts in the comments section below.