An integrated organisation can deliver an integrated service. But an integrated service is unlikely to be able to deliver an integrated organisation.

That though is a distinction which is easily missed, not least in thinking about joining up government services, where the hoped for magic of an integrated web platform leading to seamless information and service delivery has unaccountably failed to happen for ten years or more. That’s not from lack of ambition or intention: part of the founding myth, first of UK Online and then of Directgov, was that seamless citizen-centred services would spring into being. But they didn’t. There are several important reasons for that. Some of them are to do with the maturity of technology, some to do with maturity of understanding of the place of web services as a dimension of overall service provision, but some are to do with a lack of alignment of roles, perceptions, and goals of the different organisations involved.

The founding myth had some substance. I remember the excitement of a workshop designing the ‘having a baby’ life episode (as they were then known) for UK Online, with people responsible for different parts of the overall service meeting each other for the very first time and seeing opportunities for more effective service design as a result. But for all their many virtues, neither UK Online nor Directgov ever did transcend the underlying structure of government on which they were built and even now it is all too easy to see departmental boundaries not far below the surface.

The difference with is not that it’s a better website, though it is. The difference is not even that its first design principle makes an unambiguous commitment to start with real user needs, though the clarity and prominence of that principle is a wholly good thing. The difference is all of that put together with a level of focused determination and leadership direction which we have not seen before.

The question is whether that is enough. I am not sure that it is. Everything GDS does reinforces my confidence that we will have a more integrated digital platform leading to more seamless information and service delivery. There will be far greater homogeneity of how government appears to the outside world:  the visual symbolism will carry a very strong message that all are parts of a very coherent and integrated whole. Government will look and feel like one thing rather than many. The circumstances in which it is important to understand which bit of government does what will be fewer. But underneath, different things will continue to happen in different ways, as a consequence of the underlying structures.

GDS is creating a representation of government and what it does. What it can’t do (and can’t be expected to do) is make the substance of government align with that representation. That matters for reasons which in effect are a slightly different version of Conway’s law: the perceived difference between a service and an interface is very strongly driven by organisational boundaries. That means that the experience of meeting needs is unlikely to be anywhere close to being as integrated as the experience of discovering information about how those needs can be met. At a more detailed level, it means that information held by or provided to different government service providers is unlikely to be managed (or manageable) in a way which supports the aggregation and integration of services.

Does any of that matter? I think it does, for two reasons. The first is to avoid the trap of assuming that because visual identity and information are much better integrated that the problem of joined up government has either been solved or gone away. The second is that if we do aspire to organising public service delivery around people and their needs, we need to get down to those deeper layers to make it happen.

To give one minor and largely inconsequential example from another sector, the smooth integration of Amazon’s website is second to none, especially bearing in mind that it is both its own shop window and a channel for the thousands of smaller providers in the Amazon market place. I can search and buy without particularly caring how Amazon – or anybody else – is organised, and without any notion of where the stuff I am buying actually is.  But behind the scenes, the consequences of my order may well require activity to be fragmented and duplicated – and the effect on me may be that several parcels arrive on several vans at several times over several days. That’s no more than minor inconvenience and inefficiency. But if the supply side fragmentation covers tax and benefits, health and education, the consequences can be much more significant.

Of course, just integrating everything into a single monolith isn’t likely to be a good response to that problem. It would be nice to get a single parcel on a single day, but that would require the existence either of a single warehouse with absolutely everything in it, or some kind of aggregation hub where orders could be assembled. The drawbacks of either would be pretty substantial.

Back in the world of government, I feel no pressing need to renew my driving licence at my local hospital, and have no expectation that teachers should be able to process benefit claims. Here too, not everything which can be integrated should be. But the lessons of the having a baby life episode still apply: if we want people’s experience of a service to have a parallel level of integration to information about that service, we are going to have to look to the base as well as the superstructure.


  1. Heya – this was not going to fit into 140.

    I think you’re talking about a required culture change to occur. Integrated service isn’t really about passports from hospitals as you say, but it is about better communications between individual sections and departments when it does become necessary for that to occur. I roll out my oft used example of two specialists inside the same hospital writing letters to each other. All it takes is one person to mislay the letters and my life suddenly becomes very complicated as two specialists flounder trying to remember who I am and why I was there. My treatment could potentially halt. Consequences.

    You’re also right, I think, in that I don’t believe chucking everyone into a single automated call centre would work (hideous), nor that a single database for everyone would work either (hideous). What I do think is needed is a single contact phone book, with keywords attached to everyone which describes who they are and what they do. A way of people quickly establishing who can help them deal with the exceptions. You’re right that the end user doesn’t care which initials are dealing with their case. They just see ‘government’. The best that government can be is to provide that seamless service and for no one to ever see how hard the ducks legs are flapping under the water.

    How do we stop the ducks legs flapping so hard? I think by offering to help where wanted or needed to unravel iterative complexity and simplify process. No one has time to do this in parallel to their day jobs so maybe some kind of hit squad is needed – a mobile group of digital/process analysts/change bods who can move around through government leaving a similar but not identical fingerprint of service delivery behind them.

Comments are closed.