I keep having conversations with government clients who are considering to either use or rationalize cloud-based services. By “rationalizing” I mean to provide definitions, frameworks, procurement vehicles, sourcing guidance, and so forth to facilitate the use of of cloud-based services by government agencies across an entire jurisdiction.
In both cases I still sense a significant degree of confusion in definitions and approaches concerning what the US National Institute of Standards and Technology (NIST) calls delivery models (infrastructure, platform, software as a service) and what it calls deployment models (public, private, hybrid, etc.).
On the vendor side, everybody has or is about to have some form of cloud offering, and they are all said to ready or “almost” ready to support government requirements. I was watching the excellent cloud shootout organized by Fedscoop (here is the video) with companies like Amazon, Google, HP, IBM, Microsoft, Intel and salesforce.com, and the tone of the discussion reminded me of the vendor panel we had in Orlando with Google, Microsoft, Terramark, salesforce.com and CSC. The message is always that vendors are ready, solutions are available (although some are still tweaking their security compliance efforts), many are already deployed (plenty of SaaS stories have turned into cloud stories, as if it was the same), and the burden is mostly on government agencies to be bold enough to take their steps into (or onto?) the cloud.
Similar messages, although more tempered and less packaged in marketing jargon, come from those who are involved in furthering a cloud computing agenda in government. These are both government officials and consultants and advisors (often directly or indirectly associated to industry) who are tasked with developing cloud computing strategies, program offices, framework programs and so forth.
I am sure everybody agrees that cloud computing as a concept is as fascinating as it is inherently difficult to frame and describe. In any industry, conversations about cloud computing are a mixture of outsourcing, anything-as-a-service, flexibility, dramatic cost reduction, commoditization, IT utility, managed services, and so forth.
In government there are additional causes for confusion:
- Government can be both a consumer and a provider of cloud-based services: many discussions about the nature of private, community or public cloud fail to recognize this fundamental ambiguity and the associated dynamics of being a user, a provider or both.
- Cloud computing interacts and overlaps with whole-of-government “shared services” and “consolidation” initiatives. It is quite customary to see a whole-of-government data center consolidation strategy be part of a cloud computing strategy. However experts seem to forget that the causes for failure of those earlier initiatives (mostly related to governance and control) may equally undermine a whole-of-government cloud computing strategy.
- Cloud computing interacts and overlaps with government 2.0, open government and social media strategies. This is both because socialization and cloud trends have emerged at the same time, but also because it is easier to associate open data to lower security requirements and public cloud.
- The budget process in most governments discourages the shift from capital to operational spending. Of course many governments are in efficiency-seeking and cost-reduction mode, but from a departmental perspective the shift to opex implies less discretionary budget, and whatever is saved out of that budget is not likely to remain with that department
The points above need to be addressed by those who develop whole-of-government cloud strategies, and must be understood by vendors, for cloud computing to really take off and deliver its modernization payload, rather than remain at the margin of niche open government initiative and new applications.