Yesterday I attended a session about cloud computing organized by the central IT organization of a US state. Like many internal government IT service providers, they are developing a cloud computing strategy to improve their service offering to client agencies and to reduce their costs. I was positively impressed as they have been focusing on very concrete system infrastructure (aka IaaS) services, piloting them before promoting them to their clients. Given the impending elections and the challenging financial situation, they do not feel that articulating a longer term strategy, e.g. looking also at software (aka SaaS) and application platform (aka PaaS) services, is a priority, although they have rather clear ideas about to address the latter too.
We discussed about how, in spite of the constraints above, they should position their current efforts as the initial steps of a roadmap leading to a more comprehensive and somewhat compelling cloud computing vision. As suggested by efforts undertaken by the US federal or the UK government, a keystone of a government cloud computing vision is the so-called “storefront”, which is meant to be the primary channel to purchase, configure and use cloud-based services.
While this IT organization is clearly focusing on a provider view (“how can we provide services that are equally or more competitive than what vendors’ ones?”), they should also take a user view and look at how to extend or even transform their role from being service providers to being service brokers, procurement partners and advisors for their clients, in order to help them identify the right solutions to meet their needs, be those cloud or not, be those internal or external.
While I was impressed with both the down-to-Earth approach and the willingness to challenge themselves, I observed (and told them) that they were exhibiting a vendor behavior at times. This is a pattern I have noticed with several, rather mature and well-run government IT organizations as they approach the opportunities and risks of cloud computing.
My suggestion to them is always to ask them the following question:
- Which of our clients do really need cloud services and why?
- For various categories of services, should we be the providers, should be manage the procurement vehicle (e.g. the storefront) or should we let our clients make their own vendor choices?
As usual, putting cloud computing in the context of other, more traditional delivery models helps formulate the answers to those questions. At the end of the day, for how sexy and transformative cloud computing may look like, it all boils down to good old sourcing.