Wasting Money

In this episode of Beneficial Intelligence, I discuss wasting money. The business always complains that IT is costing too much. That is because we are wasting so much money. We're on track for worldwide IT spending of about 4 trillion, and surveys show that at least 25% of that is wasted. That's one trillion dollars we waste. 

IT organizations waste money in two ways: With what we build, and with what we run. 

It's a time-honored tradition in IT to never retire an old system. We keep everything running forever, even though the business benefit has long since evaporated. We might suspect that a specific system is not providing business benefit. But we don't know if anybody might be using the system for something. So we keep it running. To address this kind of waste, have someone examine the system logs to find who uses the system. If you have an old system without proper instrumentation, you might need have your network people to look at the network traffic to identify unused or barely-used systems.  

But old systems is only one way we waste money on the systems we are running. The new and much more efficient way to lose money running systems is to move something to the cloud. In the history of cloud computing, there has never ever once been a case where the cloud resource cost was less than anticipated. More often it is 50%, 100% or 200% larger than expected. Cloud pricing will charge you for CPU cycles, storage, data traffic, messages passed, API calls, and a million other things. Nobody is able to calculate all the costs, and there are always surprises. Your business case for moving something to the cloud needs to show a dramatic cost saving in order to justify the move. 

It's bad enough we lose a lot of money running things. But the money we waste running things pales in comparison with the money we waste building things. If you are still running waterfall projects, you know that they will never build what the customer wants. Many agile IT organizations don't fare much better, because the business can't be bothered to pay attention to the biweekly demonstrations of new features. Only the few organizations where both IT and business is agile don't waste money building something nobody needs. 

I've saved the largest money-waster for last: Unnecessarily building your own. IT organizations want to build systems. They don't want to buy systems. The reason is that the business always wants something changed, and if you have all of the code in your own repository, you can change anything you like. If you have a standard system, you will sometimes have to tell the business that something cannot be changed. Because of this, and because we don't calculate the business benefit of IT systems, we default to always building our own. 

Building your own is correct for systems of differentiation and innovation where you can do things differently from your competition. But building your own accounting system because you want to accomodate a very specific kind of invoicing you have in your business will never pay off. 

You can save some money by finding and retiring under-utilized systems. You can some some more by being very careful about doing lift-and-shift of existing systems to the cloud. You can save some more by getting the business on board with Agile. And you can really save money by not building your own. 

 ------ 

Beneficial Intelligence is a weekly podcast with stories and pragmatic advice for CIOs, CTOs, and other IT leaders. To get in touch, please contact me at sten@vesterli.com