This extraordinary talk by Mike Bracken (Exec Director of Digital in the UK Cabinet Office) at the recent Code For America Summit is rich with insight into transforming digital service delivery and builds on his excellent Strategy is Delivery post that I wrote about here. The interesting thing about GDS in this context I think, is that it is not only a great examplar for other governments and public sector organisations, but that there are so many learnings that can be applied within large businesses as well.
For too long, says MIke, the debate has been about policy and yet what we're living through is not a policy option, but a delivery crisis (government is full of empty strategies, delivery is the driver that gets stuff done). Hundreds of thousands of people use government services everyday, and yet the needs of government has traditionally come before those of the users. Lesson: avoid an over-emphasis on strategy at the expense of execution, experimentation, user need or experience.
GDS is part of the Cabinet Office. They are a small team that sits at the centre but has far reaching powers and works at scale - 60m population, 4.8m businesses, 1.3 Bn transactions a year, 24 central govt depts...a single domain with 7m visitors a week, 600 publishers, 5,000 more coming soon, 2,000 code releases in their first year. Lesson: a small empowered team sat at the centre of an organisation can do big things.
Identity is a problem no-one has really cracked, so they federated that problem, riding on the back of others who are already good at it (banks, telecoms and so on), presenting users with a choice who validates your identity, switching it from something the government does to something users can select. Lesson: don't be afraid to partner in the right circumstances
The approach to technology is just as interesting. Most government services have traditionally been served by a bewildering complexity of technologies. It is, says Mike, a systemic problem that we see technology as primary, and digital output as secondary: 'when we put digital in technology, we are starting in the wrong place, and guaranteeing failure'. The problem with digital services being seen as the outcome of technology purchases is that the organising principle is procurement, whereas for digital it is the user (interestingly, he gives the example of the bloated budget for Healthcare.gov being central to its failure). So the CTO for government works into Mike. Rather than IT determining the digital service of the future, it becomes the thing you go to when you need some technology. Technology is critical, but spend and structures get in the way (and tech needs to get out of the way). It is a fourth order question:
- User need
- Policy need
- Operational need
- What technology might be needed to provide the service?
Taking a different approach to IT has led to first year savings of £500m. Lesson: digital is primary, technology is important but secondary.
Mike then goes on to talk about the importance of measurement, transparency in performance, and realtime metrics (live service data) and using measures to secure support, and provide proof of benefit. Inevitably with a relatively small team, prioritisation of effort is key, so this became about focusing on the top 25 services that drive the most important and the highest quantity of transactions and running a programme to transform those ('rather than go for stuff that is neat and cool, go for stuff that makes a big difference'). Of the 660 services that government provides, 90% are covered by that top 25, and 97% by the top 50, but focusing like this enables them to make the biggest difference in the shortest time. Lesson: prioritise the stuff that will make the biggest impact to end users
Money, unhappy users, over zealous security (the 'pernicious view that security should come before usability at all times'), and procurement represent a 'square of despair'. So to sum up, digital offers a way to route around and ultimately to transform:
- Work on stuff that matters:- make services that matter a great deal to a lot of people but also ones that only the government can do and should do. With finite resources, if someone else can do it, let them.
- Deal with the enterprise:- procurement processes need reform (80% of the IT spend is on 7/8 companies so they have set up access to a far wider group of suppliers). Don't play on the kind of terms that lead to huge inflexible requirement documents.
- Get the best talent and start recruiting now:- think about creating the kind of culture where the best talent wants to work
- Start reforms the hard way:- make the infrastructure work for you
- Delivery is kryptonite
If, says Mike, we don't deliver what users need, in a digital world they'll simply route around us. This is not about changing government websites, this is about changing government.
Worth watching the whole talk. Excellent stuff.