First, let's start with a key idea: Cloud is a service, not unlike Valet parking. I'm a self-parker at heart, but I recently felt compelled to try the Valet parking. It provided me a value- the ability to simply punch a number into my phone and find my car waiting for me when I was ready for it. Now, this value came at acost, which would surprise no one - we fundamentally understand that a service, which provides added value, comes at a cost. Yet I still hear confusion surrounding the cost of cloud, and specifically some who believe cloud is supposed to be less expensive than traditional on-premise solutions. If cloud is a service that provides a value above what I currently have, it stands to reason that it comes at a cost above what I am currently paying.
Now, cloud advocates want to disagree here because the cloud service is, in theory, saving the organization resources. For example, if you are fully in the cloud, you don't need data center gear or the power/cooling required to keep it online. These are hard cost savings that help justify a cloud expense.
A cloud service also saves the organization time. Every aspect of firmware management, troubleshooting network problems, and reacting to hardware failures is now in the hands of your service provider. Time is a soft cost, it doesn't save on the bottom line, but it makes the IT staff more efficient.
I'll tell you that, from my experience, the hard costs are never able to surmount the monthly expenditure of a cloud subscription. In fact, I am part of more conversations about cloud retreat than I am about cloud migration, and I believe this is because some organizations leapt fully into the cloud with the understanding that it would ultimately save money. The problem is that the hard cost reduction didn't make up for the cloud spend, and it's difficult to evaluate those soft costs of time and efficiency to determine whether it was truly a positive business outcome.
The Cloud Value
So is cloud a bad move? Hardly! Believe it or not, I'm a major advocate for cloud. Or, more specifically, for the proper leveraging of cloud resources. Cloud should not be leveraged to save money, but to improve processes and efficiencies where it makes sense to do so. As such, here are 3 ways I would recommending utilizing the cloud:
Websites and public/DMZ resources - they're supposed to face the Internet, so let them live out there!
Spiking workloads - those that require exorbitant spend to deal with recurring "busy seasons"
Application-based moves to SaaS providers, such as Office365 and Student Information Systems
There are certainly more use-cases, but beware the temptation to turn off your data centers, because a full cloud migration is almost never the right answer. Instead, think through your applications and design a hybrid cloud model that is right for you. And hey, maybe even call your local cloud experts. It couldn't hurt to throw some ideas on that whiteboard.