The SaaS and public cloud provider model is an inevitable direction for IT infrastructure, but I also believe the market drivers moving organizations to the cloud are incorrectly portrayed; or at least slightly skewed. Cost savings and ease of implementation are very real for start-up and small IT organizations, but I’m suggesting that very few large organizations are realizing a level of cost savings to justify the move.
I say this because a real driver to the cloud is an opportunity for development teams to circumvent dysfunctional infrastructure organizations. There is a belief [and a reality] that infrastructure teams are a bottleneck to enabling business strategy. They slow down the speed to market and they disrupt the agile development delivery cycle. If circumventing infrastructure teams is an actual market driver for using cloud services, then the IT organization as a whole is not focused on maximizing the cloud strategy and will not capitalize on potential cost savings. How could they when the strategic driver is to circumvent infrastructure, a huge piece of the organization?
Historically the IT departments (Dev and Inf/Ops) had to work together out of necessity. However, the public cloud has created a new option for dev departments. It’s an option that allows them to swipe a credit card and “get” their own server without help from any other team. The larger an IT organization is, the easier it is for one obscure team to do their own thing in the cloud. However, this approach is short sighted and ignores the amazingly difficult life cycle management of infrastructure services; regardless of them being in a private data center or a public cloud.