Thoughts for GDS SPRINT15 event

Screen Shot 2015-02-04 at 16.46.53

Here are some of the thoughts I shared with the Government Digital Service Sprint 15 audience on 3 February. I had been asked to comment on what we who work in the public sector can learn from looking across at other digital business models that have emerged elsewhere:

Let’s start with a definition of ‘digital business model’: a model that exploits shared technology & process. Another, more detailed definition is that “digital is a participatory layer of media that allows users to self-select their own experiences, and affords the organization the ability to bridge media, gain feedback, iterate their service, and build & activate relationships”. Sounds familiar?

Well, the business models are familiar, too – because they’re the most discussed ones around: P2P lending; MPesa; Spotify; Skyscanner; eBay; Airb&b; Tripadvisor; Rightmove; Uber; Amazon; Click & collect; Clubcard; Black box car insurance; global supply chain integrators like Walmart. They all broker new, direct, relationships between people, services & things – & learn as they go by harvesting data. Importantly, all have been vigorously resisted by incumbents!

Now, these principles of course hold revolutionary promise when translated into the pubsec. Digital platforms could support connected cities, biometrically-supported care packages, innovations around public data, simplified, mobile interfaces, and services we haven’t even thought of yet. Tim O’Reilly’s ‘vending machine govt’ – the current setup where you pay your tax and choose from a limited set of prepackaged interactions with the state – could be gone for ever.

So much for the shiny future promise of digital: but there’s a hard lesson here for the present. Looking across at these other examples of digital business, we can quickly see that all share common architectural features. They are brokers, not providers; disintermediators, disaggregators, and disruptors; horizontal, not vertical, and movers from ‘push’ to ‘pull’. They didn’t try and bolt on shiny tech to old, incumbent business models. Much as Netflix replaced Blockbuster, digital businesses are founded on the basic understanding that the analogue way of doing things in a particular sector has had its day. As we and others have pointed out in recent books, our public sector service model is similarly bust – because we can’t afford its creaky plumbing anymore, and it can’t give us what we now want. For the cost of the £10 billion Aspire contract, a back-of-the-envelope calculation suggests we could have bought 40m GP visits at £25ea. It’s the back end, the plumbing, not the front end that’s holding us all back.

For there are two, unpalatable facts about digital business. The first unpalatable fact is that an agile, citizen-centred, data-driven ‘learning state’ requires better, not less, architecture; modular, digital business is architected well. ‘Tight-loose’, as Francis said several years ago. ‘Tight-loose’ is hard to achieve though, because working out what needs to be ‘tight’ requires serious clarity of thought: cleaving to the signal of open standards amid all that fashionable noise of agile. Digital requires ‘inside-out’, not ‘outside-in’ architecture: just as the deregulated electricity ecosystem of generators, grid and suppliers all came together around an MPAN centred on the consumer – so the ecosystem architecture of digital govt can only really come together around personal data stores. So, Architecture.

The second unpalatable fact about digital business is that digital business models entail digital politics. For they expose the legions of workers in vertically integrated bureaucracies – private and public sector, the distinction’s outdated – that effectively extract rent from controlling, and brokering, transactions between citizens. Just as in every other digital business, achieving digital government therefore involves the politics of a pronounced shift in resources – away from predicting, specifying, commissioning, modeling, building, measuring, planning, managing, auditing, and reporting on services to – well, serving on the front line, supported by a minimum, utility capability set that is consumed like power.

Here’s an example of what I mean: the Buurtzorg community nursing organisation in Holland, where a back office of 30 supports 7,000 frontline nurses; no middle mgmt at all. An audit firm reported an ‘astonishing’ 40% less hours of care for the same clinical outcome. They estimate that Holland could achieve savings of 2 billion euro each year if all their nurses achieved similar results. Grossed up for the UK, that’s almost £6bn saved, every year. Just in community nursing. Viewed this way, public service cuts start to appear as a myth perpetuated by a public sector reluctant to reform its own business model, and a collective political class incapable or unwilling to explain to the electorate what this means in practice: we could genuinely have more public services, provided our public sector reconfigures itself away from bureaucracy and back again towards serving the public.

So in summary: looking across at other digital business models, digital’s certainly all about giving users what they want, via the interface they want, when they want – and for less; but the really exciting part, for me, is that if properly architected and astutely politically stewarded, digital could entail a wholesale reimagination of the way we serve each other face to face, and actually increase, rather than cut, our public services.

It’s just unfortunate that this is a line you won’t hear during the election.

Would you like to comment?