Painting the Clouds


© 2008 Colin Currie

EDUCAUSE Review, vol. 43, no. 6 (November/December 2008)

Painting the Clouds

Colin Currie

Colin Currie is Director of Administrative Information Services at Princeton University.

Comments on this article can be sent to the author at and/or can be posted to the web via the link at the bottom of this page.

One of the great ironies for those of us in higher education information technology is that, in the coming years, doing the best-possible job for our institutions will mean finding the optimal ways to replace our functions with outside services. In other words, a critical part of our job will be to outsource ourselves as effectively as possible.

The reasons behind this new job responsibility are several developments that people may already be weary (and perhaps wary) of reading about: cloud computing, outsourcing, software as a service (with the cute acronym SaaS, which sounds oddly Swedish to me). Sometimes writers will give that usual list and finish it off with “and other things,” without venturing a guess as to what the other things might be. In any case, what this comes down to is that much of what the current IT organization does for its institution will be performed by external entities in the future.

Before panic sets in, we should realize that these changes will not happen tomorrow. As well, when we step back and think about the situation, we might see that this is all good and healthy. The IT organization barely existed forty years ago, was something people were becoming vaguely aware of thirty years ago, and was considered a necessity twenty years ago. In all those years, the IT organization has played an essential role in shepherding the new tool of computer/information technology into academia. Those of us in the IT organization have therefore been critical contributors to a watershed event in higher education. Information technology has already directly affected the core of higher education: from what is taught to how it is taught. Information technology has made new forms of education possible, and we are just getting out of the gate. Technology is maturing, expanding, standardizing, and commoditizing. But what was once a trail being blazed by pioneers has now become a superhighway shuttling family vans on cruise control into the future. The role of the internal pioneer is transforming to that of a scout who is looking for the best routes, for the best external solutions.

The idea of shifting the administration of IT functions to external entities is not a new concept. In the five years since Nicholas Carr wrote “IT Doesn’t Matter,” the wave his article created is still rolling along and is yet to crest. The different realities of each of our institutions will determine the priorities and the pace of change in our own environments, but we are all sure to be affected by the coming change in some way. We should bear in mind, though, that Carr’s article was not aimed at higher education information technology specifically. He was discussing information technology in general. This means that since we in higher education are likely to be behind the commercial sector in this change, we can look there for a preview of what’s coming our way.

The paradox of all of this is that as the IT organization moves to this new model, our jobs will get both easier and much more complicated. The parts that will get easier are the nut-and-bolts of the functions that will be moved off campus: configuring machines, patching, backing up, and so on. The parts that will become more complicated relate to the increasingly distributed nature of how we will fulfill technology needs. Some vendors will want to do it all, but many schools will likely spread externalized services across multiple providers based on the fit of the solution, perceptions of the service, and the projected costs. One area that will surely become more complicated, for example, is managing and controlling IT security. With data residing in the cloud, with externally hosted applications authenticating campus users, and with IDs and passwords residing who-knows-where, our already problematical efforts to stay a step ahead of hackers and identity thieves will become even more complex.

Keeping all of this straight and working optimally will require good management skills and also—above all else—a functioning project-governance structure on campus. Indeed, project governance is the factor that will separate the institutions that will thrive in this new reality from those that will struggle. Because many traditional IT services will no longer be under the direct control of the IT organization, the connections and communications between the IT organization and its customers will need to be the best they’ve ever been. Also, because the IT organization will be a critical participant in the decision-making about what, when, and how to shift services to third parties, those in the IT organization will need a clear and comprehensive understanding of all of the technology initiatives across the entire campus in order to establish a service mix that will work successfully.

We have some very important advising to do. Our institutions will need help to understand what parts of the current IT functions can be provided just as effectively—or (shudder) provided even better—by external entities and when the time to make the shift has come. We should be researching the best alternatives to having our organizations perform specific functions. And we should be making honest, critical assessments of when our systems and processes will hit the tipping point where it makes sense to have someone else, from somewhere else, provide the service.

Perhaps we can move one step further, not only identifying which functions and when, but also defining what makes the externalizing of IT services a success overall. By coming together and speaking in a consistent, unified voice, we can help identify the data standards, protocols, and interfaces that will result in the best-possible external service offerings. We can help to define, for example, what applications we would most like to see coming from hosted sources, what data elements they should contain, and how they should look and behave. Once some of these key elements are specified, we can then share this information with vendors and other open-source entities. In that way, rather than having these major and critical parts of our future designed for us, we ourselves can help to paint the clouds.