Picture credit: @kitterati from Sprint 16

Setting standards that stick

Use them to empower the disempowered.

GDS launched the digital by default service standard and accompanying service manual in 2013. The objective was to make all government digital services so good that people would prefer to carry out the task online. Every new service launching on GOV.UK had to pass. As product manager, it was the most important 523 words I worked on in government.

Standards and manuals have become popular. Australia and Scotland have adopted their own pithier versions, The U.S. Digital Service and 18F have built a playbook. Even the University of St Andrews has one. Smart people like Carl Haggerty and Dan Slee in local government are taking a closer look at what they should do. GDS is thinking about how the standard could evolve.

When we started working on the standard, we thought the hard bit would be working out what should be included — where should the bar be set. That’s easier than you’d think. There are plenty of people working in an organisation who know exactly what needs to be done. They will tell you if you listen. What they lack is the support to make change happen.

The responsibility for those with influence is to find the people who want to do the right thing, and remove everything that gets in their way.

Put powers first

Big organisations run on inertia. That doesn’t mean they stand still. They resist change in their direction and speed. If you’re trying to set standards for what good looks like, to have a real impact, you have to deliver something that disrupts inertia. So to have real bite, you need standards attached to powers. Look forward to becoming unpopular.

The first goal for any set of standards is to prevent terrible things from happening by default. That means stopping things that may have lots of momentum, money and egos attached. Everyone knows these projects; juggernauts careering towards the cliff edge, or zombies that just won’t die. Everyone knows they will fail. Everyone feels powerless to stop them.

Faced with impending disasters, writing down a list of things you would like to see and keeping your fingers crossed is a waste of time. It won’t stop them. Dusty piles of optional guidelines and best practice manuals coexisting with manifest disasters are testament to this.

And if you don’t have powers — like the ability to stop spending, or prevent something being launched — forget about standards for now and concentrate on building your credibility by delivering things that work first. Earn the right to be trusted as a good judge before appointing yourself.

Be so open it becomes boring

Once you’ve got some powers, it still takes confidence to stand up and say ‘we’re not doing it that way anymore.’ You’ll need to bring others with you. Having supporters is essential, particularly as those who have done well out of the previous arrangements begin to realise you’re not going away.

For people to back you, they’ll need to be assured you’re right and that you’ve listened to them. Show all your working to everyone who wants to see it. Be humble about mistakes you make, don’t apologise for the fact you’re learning as you go.

Accompany the stick with the best carrot you can provide, and give it away to everyone for free. The service standard would have failed without the design manual. The manual built goodwill, clarified meaning and had collective ownership. Keeping it fresh was a cost — but it’s one paid for hundred-fold by avoiding service failures and attracting good people to come and work with you.

And be upfront with your senior sponsors. Tell them you’ll need them when shizz gets real.

Fail with style

Because shizz will get real. Sooner or later, you’ll be confronted with a juggernaut or a zombie, and your shiny new standard will be in the spotlight.

When it comes in, follow your process properly. Don’t make it a special case. Have the right people in the room. Be unimpeachable.

And if it isn’t good enough, say so. Be human about it. These may be things people have worked on for years. Don’t be high-handed. Don’t crow. Don’t patronise. Say no with class. Cash in that senior support (‘I’ll sit next to you on the Public Accounts Committee to explain this if you want.’)

It’ll be painful, it’ll take a long time — but it’s an essential line in the sand. You are saying politely: this is no longer acceptable. Remember that if you don’t, if you capitulate, then your standard has failed at the first hurdle. Inertia will barrel on.

When the dust settles, word will spread. You will be called jobsworths, meddlers, obstructors. Don’t go round looking for credit. The people who know what needs to be done will be silently applauding you.

Build the tiniest possible thing

My big regret about the service standard is that it became a monster. @jukesie, as ever, summed it up nicely. This wasn’t our aspiration, but there may have been nothing we could do about it.

The first assessment against the standard took 25 minutes, covered three points, involved four people and resulted in one email.

These days, a typical assessment against the standard takes 240 minutes, covers 18 points, involves 8–10 people and results in a long report and lengthy email exchange.

My point isn’t that it’s all gone awry. It’s that nothing ever gets taken away from processes. They accrete and grow. Every small addition is a completely sensible step. But then you step back and realise a behemoth has appeared. If you get the chance to start afresh, start small.

So if you’re in local government, how do you do that? Start with a blank sheet. Go back to the people who want to do the right thing but can’t. It will be a different group. Ask them what they need. In 2013, the response was user-led design, agile working and open source software. In 2016, it might be end-to-end service design, registers and knowledge transfer to public servants. I don’t know. Your users will.

Standards don’t need to be comprehensive. We didn’t write anything about business cases in the service standard because inertia pulls in that direction anyway. Go for what’s not done. Shift the Overton window. Use standards to empower the disempowered. And make them stick.

@ad_greenway