Platform thinking

As part of my work for central government, over the past few weeks I’ve been immersing myself in the work of Tim O’Reilly and the concepts of Government as a Platform.

The concept at its heart is a metaphor. In fact a metaphor of a metaphor which is always a slightly risky approach prone to misinterpretation: a platform is something on which one stands; a platform in computer and technology terms is a set of stuff on which other things can be built – the IBM PC; Microsoft Windows; the Internet; the World Wide Web; iOS; Cloud computing services like Amazon Web Services; commercial web services like Amazon, Facebook… The list and the evolutions go on.

Government as a Platform takes from the experiences of the world of technology and posits that maybe Government that is fit for purpose in the 21st Century should be more like a platform (providing services upon which others can build), and less like a vending machine (you put your money in, you get your services out).

O’Reilly extrapolates three key lessons to be learned from the world of tech that should shape the design and development of Government as a Platform:

1. open standards spark innovation and growth

Now there is a difference here between “open” as in shared and “open” as in commons; it’s one thing to follow Open Source standards for everything, but proprietary yet open platforms can be just as successful as ones that stem from the stable of GNU. Witness Microsoft’s empire, Facebook, the retailing power of Amazon… Conversely, innovation and growth won’t come just because you’ve decided to adopt Open Document format.

This ideal provides some interesting challenges for most organisations, not least government. Most traditional hierarchical organisations (ie most big ones) are closed by design. We do things behind closed doors; we protect our information and services; we control and tightly manage our interfaces to the outside world. That world is gone: the means of mass publication and collaboration are available to anyone with a smart phone. But dragging big, closed-thinking organisations into this new world is going to involve a lot more kicking and screaming than we’ve seen to date.

2. build a simple system and let it evolve

This is basically the agile story. If you look at the big successful things that surround us in our modern world, they’ve all evolved. None of them were blueprint-planned. They kind of happened. Serendipity, lucky choices, right places, right time, gambles, hunches. Anyone who thinks otherwise has been supping too long on the Kool Aid of Big Data and will be soon to feel the consequences.

However, once you become “big” the rules change. If you are a big publicly-listed company, you become beholden to the Gods of The Investors. Their Words become True and you shall follow Their Bidding. Which basically means too often an organisation becomes structurally unable to say “We don’t know but we’re going to try” because you’re not allowed to say “We don’t know” ever if you have (particularly activist) investors.

In the public sector, if you are big, you are beholden to politicians and the press and they like a nice scandal to dig their teeth into . “Don’t know” becomes an unacceptable viewpoint (even if everyone knows that that’s the only sensible viewpoint to hold).

In a world absent of “don’t know”, simple and evolving systems become big, ungainly, centrally planned failure. But moving to a truly agile approach involves changing the mindsets of all the naysayers, or finding an alternative delivery model. For the private sector, particularly in tech, that alternative model increasingly seems to be the acquisition of what look to be high(er) potential start-ups, where some of the early failure has already occurred elsewhere and on someone else’s balance sheet. It’s going to be interesting in government as to how similar models could evolve: in the old days this was called nationalisation, but apparently that’s only acceptable these days if the acquisitions are on death’s door and too big to fail.

3. design for participation

This is the lesson of the open source software movement; that both products and development methods in the open source world have by necessity been designed for multiple parties to contribute and engage.

This is an interesting one in the public sector: engagement in traditional political mechanisms is at an all time low;  developing new models for engagement might unlock that. But conversely there is a risk that in a world where people can’t be bothered to be involved, kick-starting them into being bothered by expecting a lot more involvement is a bit of a self-referential loop. It shouldn’t be ignored that the old world of “vending-machine” service delivery involves much less effort for the end user than government as a platform.

It’s a bit like the difference between final salary pensions and defined contribution models that are now more common. In the old world it was easy: pay my subs, get my pension. That it was utterly unsustainable is difficult for me as an individual to countenance. The new world looks rubbish, where suddenly I have to take much more responsibility for my own outcomes, and take on much more of the risk myself.

Government as a Platform is an interesting metaphor. There are precedents: O’Reilly points to the roads network as a case in point (and the Internet itself was also a state-funded platform in its early incarnations). It’s not going to be easy to achieve, not least because it faces up to hard realities about participation that many won’t want to face. It will probably be misinterpreted by many, so we’ll end up with internet-enabled Government as a Service in places as a rather lack-lustre alternative (a bit akin to how some organisations are misinterpreting The Internet of Things). But it makes some bold steps intellectually that are as much about the sustainability of all big organisations in 2015 and beyond as it is about just the public sector.