Why AWS’ Open Government Platform could revolutionise government innovation

The launch of AWS’ Open Government platform has the potential to revolutionise public sector innovation. It can accelerate the sharing and re-use of small components, loosely joined whilst avoiding existential dead-ends about governance.There’s been much talk of ‘government as a platform’ — the idea of small independent building blocks that can be joined together to build a robust, resilient digital service more quickly and at lower cost. The Coronavirus pandemic has seen lots of organisations use ‘platform’ services such as GOV.UK Notify and both the GDS and NHS service design patterns to create scaleable services in days. These also increase the quality of the user experience. Skilled developers remain hard to recruit and retain and re-using these components supports a more accessible and consistent interface for users.

Local authorities have responded to the steer provided by central government by creating 350 or so marginally different services to provide food, prescriptions and befriending to people most at risk of the virus. Whilst casual observers will return to tested arguments about waste and duplication of effort, more thoughtful observers will welcome the re-use of design patterns and ways of working whilst recognising that the needs and capacity of residents, the local authority and health partners necessarily vary between Hackney and Hull. Mimicking form design, sharing insights over Slack and case studies at a webinar enables pace and collaboration in a way that few shared services functions have achieved. Hackney’s Find Support Services map is our first listing in AWS’ platform.

Until now, ‘government as a platform’ has repeated a major strategic error tested by technology greats, whether Apple in the 1980s or Nokia in the 2000s: it’s been a closed platform. The central government body (whether GDS in the UK, USDS) has kept tight control over the platform. Only its creations qualify. Indeed, creations by other government departments appear not to qualify. The UK government platform has no mapping capability, for example.

In this context AWS’ Open Government platform could be truly disruptive: an international effort to showcase the best of government digital provision. No longer will Australia have to build its own version of GOV.UK Notify. Providing access to global products can also accelerate genuine intervention (contact tracing, anyone?)

To succeed, AWS will need to consider three factors. GOV.UK Notify works because it’s a service, not just lines of code. It’s got a heritage in supporting less confident organisations and has done years of sales and marketing. The forces of conservatism no longer mumble about whether it’s still going to be there in three years time.

Secondly, it needs to be patient and entrepreneurial in supporting organisations with very different skillsets. Most local authorities can’t simply re-use the GDS design system because they can’t hire developers who understand the code. In Buckinghamshire, just 35 miles from London, I couldn’t even persuade three digital agencies to co-locate with us to draw on their skills. Notify took too long to take-off.

Finally, the components need to remain just small enough to be widely applicable. The market is saturated with low code providers who promise the ability to produce ‘all the forms’ after a few days training. But no one has ever successfully mastered data or migrated to the nearest competitor with similar ease. If we could re-use a GOV Refer service from San Francisco or Singapore we would. But a complete case management system would be too cumbersome to customise. The role of understanding how user needs vary place by place will become more important. That’s why I’m most proud of the user research library we developed. Now, the collaboration on user research that Hackney did with Croydon to compare and contrast needs from adult social care is a further step towards adding value at the visible end of the value chain.

Some will fear creeping ‘surveillance capitalism’ or just plain old vendor lock-in by another means. Without those fears, there wouldn’t be any healthy tension in the system. I’d argue it’s precisely because the vast majority of government isn’t running on AWS (in the way it runs on Microsoft, Oracle or poor old Civica) that this risk is less than some will present. And as long as it remains an Open Government platform, it’ll be worth supporting. Our new digital services are hosted in AWS by default, but we’ve used containerisation to support portability should there be a clear advantage in another cloud platform.

The hard thing about innovation should be about building a great service that solves problems for people. AWS’ platform gives the public sector a chance to succeed.

*I previously, wrongly, asserted that the Canadian Digital Service ran a closed platform. Thanks to Ross Ferguson and his colleagues for pioneering an open approach