Government Cloud Computing - Public or Private

One of the hottest topics in government at the moment is usage of the cloud, and this was one of the topics up for discussion while I was in Japan recently.  The Japanese Government announced earlier in the year the plans to build a private cloud for Government, the Kasumigaseki Cloud.

Likewise in the UK I have working with teams on the proposed UK G-Cloud, which would be a private cloud based in the UK providing services for Government.

There has also been a lot to discussion in the US around the governments use of cloud, with the launch of that allows government organisations to purchase public cloud services from a centralised web site.

From these initiatives a number of questions and thoughts spring to mind about how the different cloud models might work in government.  Although cloud technologies have been with us for a while, how the public sector can take best advantage of them is still to be decided.

Public Clouds

There are many obvious security and data sovereignty arguments that restrict the use of public cloud services.  Governments around the world have many rules that restrict where data is located and where data can be accessed from and by who.  I think that the US Administration has been able to adopt this model in a way that many other Governments will struggle as many of the Cloud providers are US companies with data centers on US soil.

There will be use of public clouds in governments, but in most cases it is going to be reserved for public data.

Private Clouds

This is what I believe that most governments will be looking at for the majority of their cloud needs.  The risk of moving to a pubic clouds outweighs the cost benefit.  However, if we can provide services to governments though private clouds we can gain many of the benefits but allow them to control the mitigation of risk.

Clouds don't sell themselves

However, please don't believe for a moment that this is an easy sell.  The decision making process in government is complex and political.  Cost is one of many factors, and it alone is not enough to sell the cloud.  The shared services initiatives in most countries have struggled, and we run a risk of positioning cloud services as shared services v2 – which will hit the same business issues as shared services v1.

Unless we can come up with a compelling business proposition for governments (which is different from a compelling commercial business proposition which most cloud services are targeting) then they will continue to buy services that they can install and control locally.