Part one of this story looked at the ways in which cloud service options are maturing and changing IT planning. Here, the question of whether improved agility or cost savings are a bigger impetus to adoption is explored.
If she had her way, University of Michigan CIO Laura Patterson would never build another physical data center at her school. "The cloud model is a significant paradigm shift; it's not just a different way of sourcing. I think early adoption to the cloud is a competitive advantage," Patterson said.
Patterson is putting money where her mouth is. She and her IT team are in the midst of a five-year technology modernization project, called NextGen Michigan. Cloud solutions are at the heart of that endeavor. At a research university, IT is not centralized, precluding the economies of scale that a consolidated IT function can reap. Providing cloud services to the various IT departments on campus is a way to save costs while not impinging on the agility having local IT operations often affords.
Do your homework on cloud security
While security concerns about cloud computing are not unfounded, CIOs must recognize that their caution may not match the needs of the business, according to analyst James Staten of Forrester Research.
"The business simply wants to get the business objective done, and if the cloud allows them to do that, they would rather have IT fix the security problems on the back end rather than impede agility," Staten said.
In these cases, erring on the side of caution and suggesting a cloud service isn't safe can be a major misstep, Staten said.
"If you're going to argue the cloud is not secure, make sure you have an account with the service in question and that you've tested to see if it could be made secure or not," Staten said. "If you haven't done that step and the competition has, that guy wins, and Lord help you if you want credibility on the same topic again." -- K.G.
"What we're trying to move to is having cloud offerings available, so rather than buying physical hardware, they can leverage those cloud offerings," Patterson said. She believes having access to cloud services will actually result in greater agility and lead to more innovation by university researchers. "It's a way of allowing the whole community to more quickly access the technologies they need."
As for whether agility is starting to trump cost-saving as a prime motivation for adopting cloud services, John Burke, principal analyst at Chicago-based Nemertes Research Group Inc., at least, was not ready to make that call. Some larger companies still say they're going to the cloud is to save money, but often the correlation doesn't pan out, Burke said. Companies want, at the very least, to achieve cost neutrality. Sometimes, there will be a chance to spend less money, but that's not the standard for all moves to the cloud.
Burke uses his own company as an example: Nemertes recently transitioned to an Infrastructure as a Service model, moving from a dedicated hosting environment to a cloud environment.
"That saved us a ton of money on our internal server plant, and there are others who will be in a similar situation," Burke said. "But it's not the case for everybody who moves to Infrastructure as a Service, and it's also not the top goal for everybody."
Indeed, IT leaders looking for a definitive, comprehensive total cost of ownership answer regarding cloud service adoption will seldom --if ever--find one, said James Staten, principal analyst with Cambridge, Mass.-based Forrester Research Inc. By the time a comprehensive total cost of ownership study is completed, it's almost always wrong, he said -- the technology moves that fast. Instead, the key is determining what tasks need to be accomplished, and whether the enterprise is more concerned with the time taken to complete the project or with the ongoing costs once it's done.
"In almost all cases, if speed matters, cloud is going to win. If it's going to be about the ongoing cost of supporting this solution, on-premises is going to win," Staten said.
But even this equation comes with qualifications. "If the application you're wanting to deploy is super-complex and you want to maintain that same complexity when you move it, you probably shouldn't move it."
Let us know what you think about the story; email Karen Goulart, features writer.
This was first published in March 2013