I can now divulge that I worked pro bono between October 2009 and July 2010 as a member of the government’s G-Cloud/Apps Store initiative.
This arose from the Government CIO’s request to Intellect (the IT, telecoms and electronics industries association of which I am a member) to help resource this strategic study. Intellect delivered, and a wide range of industry experts were put to work – many, it should be noted, from companies who had commercial interests to protect.
As an independent advisor I believe I was able to take a broader strategic view.
Although our findings were published this spring, follow-through in Whitehall has been limited. In several of my recent columns, in the form of letters to a fictitious departmental CIO called Tim, I sought to articulate key issues.
The Government ICT Strategy has now been published, Whitehall is moving into a higher gear, and Tim has moved to a private-sector role.
The real life Government CIO, John Suffolk, has also moved on. In his final months in office, I took advantage of a webinar I chaired with John, David Wilde (Westminster Council’s CIO) and the aptly named Andy Macleod (Head of Strategy and Policy at Cisco UK) to test two particular issues.
The first lies at the core of the development of the G-Cloud itself. Is this to be a government-only private cloud, specially built and hidden behind heavily defensive firewalls, or could the government set out to make major use of the competitive economics of fast-developing public clouds?
I believe this second option makes better strategic sense. With an effective exploitation of contemporary security processes and systems, government requirements up to and including security Incident Level (IL) 3 could be serviced from large-scale multi-tenanted data centres designed to simultaneously service public-, private- and consumer-sector business. This could encompass some three quarters or more of the government’s requirements.
As an increasing volume of government/citizen transactions are done over the web, there is a certain logic in servicing such requirements from facilities already designed with the consumer marketplace in mind and already significantly cloud-enabled. IL1-IL3 privacy and security requirements are in practice much the same in both government and private sectors.
So the essence of the first issue is this: will the government seek the best service economics by taking advantage of the private sector’s fast expanding investment in multi-tenanted public clouds rather than pursuing the significantly more expensive bespoking of special private clouds for the bulk of its requirements?
The second issue lies at the core of the so-called Apps Store concept. Where apps are significantly standard in their requirements across public and private sectors (such as most back office and desktop services) the argument has to favour direct government sourcing of existing services from the competitive open marketplace.
Where the apps are more bespoke in their requirements, the real focus has to be on the implementation of more agile means of software and systems development to create and deliver them.
Actual client requirements are generally fluid and develop with experience, and aspects of enabling technologies also change with time. The classic locked down, multi-year contract is a proven recipe for failure.
So the essence of the second issue is this: will the government source those apps that are demonstrably standard from the open marketplace, while developing fresh contractual means for the creation and delivery of those apps whose requirements are more bespoke.
A study recently published by the Institute for Government called System Error – Fixing the Flaws in Government IT is worth a careful read. It builds its argument for change around the two concepts of ‘platform’ and ‘agile’.
- By platform, it refers to a shared, government-wide approach to simplifying elements of IT, “bearing down on costs, reducing duplication, establishing shared standards to enable commodity procurement, coordinating delivery of common IT facilities and services, and setting common and open standards to support interoperability”. This, I would argue, is where the government should position to source the vast majority of its requirements directly from public clouds.
- By agile, the authors recognise the need to be much more flexible, responsive to change and innovative where the development of more bespoke apps is the challenge.They recognise the transformation in government IT contracting this will require.
These two core issues should shape the strategic debate. The challenge is in the complex and parallel transformations required in the behaviours of both the UK government and its suppliers.
Richard Sykes was vice president of IT at ICI in the 1990s and is now a consultant