I had been challenging a major software developer over the last few years as to how they were going to deal with moving their services to some sort of ‘cloud’ solution. I’d then heard from a major hardware and ‘cloud’ provider that they were holding conversations with a software company that sounded very much like the one I’d been baiting. Having, as one of my last actions, pointed out the many errors in their ways to the software house before I left my job, I got a response to the question of ‘cloud’-based applications, which was interesting since I am preparing an academic paper on ‘clouds’.
The senior person at the supplier told me that they had been making pretty detailed investigations into hosting their applications as a cloud platform. One key outcome is that it became most viable if they hosted them, rather than having them hosted by a third-party – that’s according to their figures. A further revelation was how they could actually sell them at a realistic price – sticking applications in the cloud takes away many of the other costs around local hosting be that electricity, cooling, hardware and software upgrading that don’t necessarily appear in an IT revenue budget, but these costs will then appear in any ‘per seat’ costs the supplier has to charge to make a profit. This takes us back to the challenges around government accounting principles dealt with in earlier posts, and which my supplier friend agreed with.
Whilst efficiencies of scale are necessarily a key saving in general government processes, when it comes down to avoiding a multiplicity of data centres, each with their own power, ‘tin’, security, networking, cooling and management, if local authorities are all running the same application for the same service, there has to be a change of view – this has to be planned for sooner rather than later…
The Great E-mancipator 