In this post, Michael Beaven (former GDS Transformation Programme Director) shares his latest thinking on how to make digital change happen faster and highlights the key issues to address before we can turn Government as a Platform into a reality.

It is nearly a year now since the term Government as a Platform (GaaP) started to surface and the enabling strategy was being put in place to make it a reality. A lot has changed in the past year in terms of the government’s digital agenda, the nature of GDS and the maturity of departmental digital capability.

When I left GDS, nearly eight months ago now, I had a bit of a rude awakening to what had been happening outside of our own intense bubble. Whilst we were forging new services and started to conceptualise platforms, the tech market was busy creating new products that answered some of these challenges. In 2011 (I know, all those years ago) we had a limited toolset and settled on an open code model. We were consuming some open source ‘packages’ (e.g. Sugar CRM), but were largely reliant on our own efforts to create transactional components.

However, it became apparent to me late last year that there was an abundance of platform offerings on the market that could be used in whole or in part. Some of these (like Cloud Foundry) have found their way into the government toolkit. We are also seeing Salesforce and Microsoft Dynamics Online appearing more in local and central government.

One of my earliest surprises when I joined Methods Digital was how similar these big public cloud platforms were to the GaaP architecture we had been proposing within Government. Needless to say, I was left a bit deflated thinking that by the time we had got to advocating the platform idea…Microsoft and Salesforce had already built the thing.

If only it was that simple though. There are still a few topics that need a really good airing. For me they are: security, openness and lock in.  There is much negative and positive spin on these topics, so I would really welcome some conclusions. We have seen two main models over the last five/ten years, with the platform model now coming into play. I have outlined my thoughts on these below:

So in my opinion none of this stuff is perfect.  We adopted open source in 2011 as it gave us speed, some freedom and the ability to share. But it meant we had to build most things and this approach has carried on into GaaP. Note that I totally support that Government should build some platforms but it shouldn’t build all of them from scratch. Also, the challenges of growing the capability inside Government remains – especially for smaller functions/departments.

Public cloud is a critical evolution, speeding change up by using pre-built capabilities and providing a support model that doesn’t require an army of full stack developers. When you combine this industrial process and back-end with open source to provide really flexible user-driven services, I think you have something special. So the sooner we can resolve some of these questions, security issues particularly, the quicker we can start making Government as a Platform happen without having to build it all from scratch.