Notes on Cloud Computing

Cloud

Despite the hype, I  find cloud computing challenging to think about and full of interesting opportunities.  I am beginning to suspect that claims that it is a game-changer are not so far fetched.

So far, our notes on cloud computing have focused on cloud offerings as alternatives for things we can do ourselves. So, for example, Jeho wrote about ODE Architect in the Cloud and I wrote about things like Formstack and Google Mail.  These are all things we either could run locally or are already running locally.   Many discussions of cloud computing focus on the pros and cons of doing things “on premise” vs “in the cloud”.  But there’s another dimension that I have recently been thinking about: cloud offerings for which there is no local alternative.

There are huge incentives for a vendor to work with a Software as a Service (SaaS) model.  To name a few

  • Every customer has the same version of the program (the only one!)
  • No need to provide different versions for different operating systems (although there are still browser compatability questions)
  • Licensing is much easier to manage (because it boils down to managing accounts) and no one can make pirate copies of the software

So it is not suprising that we are seeing vendors starting to offer SaaS only options. I am convinced that we will see more, and that this requires a central IT organization like CIS to develop some new skills.

Here’s an example.  The HMC Office of Admission was notified last Spring that the College Board is terminating its Recruitment Plus software.  This application is used by Admission to manage the process of finding students, taking applications and making admission offers.  So it is vital to the College.  Other vendors are all attempting to get the business of soon to be former Recruitment Plus users.   An offering that our Office of Admission is interested in is delivered in a SaaS only model by Admissions Lab.  So clearly there’s nothing for central IT to do, right?  No servers to install, no software to test, no support resources to provide.  As it happens, there was plenty for CIS to do, at two levels.   At a technical level, the output of the Admissions Lab software will still need to be fed into CX, so the technical folk needed to take a look at integration options.  And, at a policy and risk management level, we sent Admissions Lab a set of questions aimed at finding out about privacy, security and data management practices at the company.  This process resulted in a much better contract for the College than would have been the case if we’d accepted the first draft.

In consultation with the Cabinet and the Computing Committee, I’ve been evolving an IT Decision making model (aka IT Governance) that will help us with these kinds of decisions. A central tenet of the model is that not all IT decisions are made by the CIO, nor should they be.  The Admissions Lab software decision is a perfect case study: it’s one where the IT decision is made by the VP for Admission, and the CIO plays a “decision support” role.  The Admission office was one of the first to make use of this model, and I thank Thyra Briggs and Peter Osgood for their patience and engagement in the process.

Watch for more news as we get this governance model developed a little better.

Leave a Reply