Skip to main content

Should the Government Develop the G-Cloud?

In the UK, we’re fairly well known for the amount of cloudy weather we have but it now seems we want more cloud – not only does it give us plenty of rain but apparently it also brings savings on a huge scale! Well that’s what the British government’s senior IT officers seem to think and they are relying on the Public Sector Network (PSN) and G-Cloud to help deliver some of the substantial £3.2 billion savings laid out in the Operational Efficiency Programme (OEP). So can the PSN and G-Cloud step up to the mark – Tom Brand, virtualisation lead at GlassHouse Technologies (UK) is not convinced.

The ‘ICT Strategy for Government’ suggests that over 1,000 data centres will have to be consolidated down to between 10-12; achievable only through huge levels of highly efficient consolidation and virtualisation. It will also mean that the vast number of government departments will have to start sharing services and infrastructure at all levels, requiring standardisation on a massive scale.

To date, centralisation and standardisation on a large scale doesn’t exactly have the best track record within the government and by the very nature of the shared infrastructure required this is unlikely to be a simple task. In commercial organisations it is easier to simply dictate strategies such as vendor selection and architectural blueprints to the entire organisation. However, across a diverse range of public sector bodies all with existing strategies and contractual obligations this is likely to involve another costly rationalisation exercise.

I’m a huge fan of cloud computing but it has to be approached correctly. Organisations that think they can treat moving to cloud computing as a simple P2V exercise need to think again. Moving existing systems ‘en mass’ to cloud IaaS (Infrastructure as a Service) solutions requires huge levels of capital investment for planning, migration and the re-platforming of existing, often legacy, applications.

It also requires significant changes in operational process to align to the highly flexibly and dynamic nature of cloud computing. If the government is looking to save £3.2 billion pounds I’m not convinced they have factored in the possibility of a £1.5 billion expense to get there.

There is no doubt that in the long run cloud solutions should be able to help organisations make huge saving by converting to an operational cost model further enhanced by pay as you use computing but it won’t be big bang!

At a recent G-Cloud CIO event I spoke to a number representatives from a range of government departments and I was surprised to hear that a significant amount of collaboration already exists across departments. A combination of this collaboration and the development of standardised SaaS (Software as a Service) solutions, similar to SalesForce, could significantly reduce the cost of developing and maintaining bespoke and over complex software solutions.

As with commercial enterprises people are coming round to the fact that an application doesn’t need all the bells and whistles; if it does a good enough job at a reduced cost and the end users like it little else matters. This all sounds great but in reality to get the benefits of SaaS it will require more than just two departments collaborating per application - something I see being a major issue. We have already seen a number of large scale IT deployments failing within single departments so how can I start to believe that multi department SaaS solutions will fare any better?

Perhaps if the government were to introduce a standardised cloud software development toolkit (similar to Google’s PaaS offering) in which applications could be developed by one department and if they are deemed a success they could simply be deployed or rented to other departments across the PSN the number of customised applications and the associated development and maintenance costs could be further reduced.

Cloud computing can easily lead to ‘your mess for less’ and without operational efficiencies and more mature service management in place the cost of running a cloud solution could easily spiral out of control in no time. However I look at it I just don’t see how the government will make the savings required by jumping on the cloud band wagon! As with any initiate like this there will be success stories but will there be enough to meet the EOP requirements?