How Alphagov might change UK government for the better

Alphagov. It’s here, it’s great, and if it’s taken forward, it will herald a great deal of positive change. Not just on the web, but in the way that government does things – and that will bring harder challenges.

There’s lots in Alphagov to be pleased about. It’s not perfect, but as an attempt to come at the problem anew – without being hampered by previous failures and hordes of stakeholders – it’s a real success. But it does beg questions that will be challenging for government to answer. Alphagov is years ahead of much of government’s culture – not to mention, most of its suppliers’ – and changing that will require leadership. Thankfully, exactly the right sort of person is in charge, which will help.

At its heart, I think Alphagov is a well-designed compromise between two design principles which are not usually reconciled very well:

  1. the desire to have a single, consistent and positive user experience; and
  2. the desire to have a maintainable and extensible platform that accommodates change and innovation without introducing excessive cost or risk.

Traditionally, government hasn’t managed either of these things well, though there has been progress on the first. The second, though, is usually sacrificed on the altar of cost savings, risk-averseness, time pressure, politics, supplier reticence and all the rest. And as a result, we end up with monolithic platforms which are expensive and bad, and support almost no innovation at all.

One of Alphagov’s main accomplishments is to avoid this trap. They’ve built a single site, but underneath it lies an assortment of APIs, scrapers, apps and bits of content. The site that you see is a clever-drawing-together of all those things in a way that gives you the impression of a single system, but which is in fact made up of many independent parts, loosely joined.

As a technical solution, this is brilliant. If you’re going to have a single platform, this is the right kind of platform to have, because it embraces change. If you want some new functionality, add an app for it. If you need a new department, add a new instance of the department app, add your content, and you’ll have 90% of what you need. If you want to run a consultation using someone’s third-party tool, just have them brand it appropriately and write an app that gives you as much integration as you want, or as the tool can support. But this kind of flexibility is powerful. In many respects, it’s anathema to the way government works.

For a start, it requires something government unwisely gave up on long ago: an in-house development team. You can no longer rely solely on content editors and web publishers to do everything. You now need developers to help do some of the things that CMSs currently do. But that’s ok. Developers can do them better, and a good in-house team will almost certainly be cheaper than most of the outsourced websites that exist at the moment.

Also, there are consequences to focusing on common user needs. Addressing 80% of users’ needs compellingly and ignoring the rest is a sound principle of design, but government can’t do that. So, if we’re to use Apha.gov for the 80% – which we should – we need another solution for the 20% of people who have strange or complicated needs. My solution would be a call centre full of wonderful, informed people who care and can make decisions. But that’s a whole other blog post.

More questions surround specialist content. A place is needed for HMRC’s tax guidance, and MoJ’s content about the courts. But those topics are not relevant to many people – they are for accountants and lawyers – so surely, Alphagov is not the right place for it.

So where should it go? And more importantly, how much should go there? Will these services be publishers of last resort, or will they be editorially controlled to minimise bloat and maximise relevance? The latter is preferable, but a publisher of last resort is necessary. In the world of Alphagov, I’m not sure there’s yet a place for it.

Yet more questions about transactions. There’s a reason there aren’t any on Alphagov: transactions are complicated, messy beasts, unavoidably bound up with business processes and legislation; empires, politics and entrenched positions; long contracts and vast sums of money.

To implement transactions following the example that Alphagov sets for everything else is a monumental challenge, because it’s not primarily a technological problem. It’s a process problem, and those are much harder to fix. Alphagov will need serious clout if it’s to make a dent in online self-assessment, online passport applications and the government gateway.

But, most importantly, these problems stand in the way of realising Alphagov’s real value at all.

Alphagov is an exercise in standards-compliant modular design. It’s built to make things interoperable, and to support diverse technical approaches. Thereby, it will make all sorts of other things possible: innovation, rapid change, a community of suppliers and developers who can be commissioned to extend the site’s functionality in various ways, or whose projects government can make use of. This is where the value of the platform really lies, and it’s what differentiates it from a well-implemented CMS. This is why Alphagov is a better solution than using WordPress.

But this approach is a million miles from what government does now. You cannot divorce the way a website works from the way the owning organisation works, and innovative, rapid work involving lots of third parties isn’t something government does well. Government is geared towards long agreements with big companies, towards rigid processes and policies, towards stability and predictability at the cost of flexibility and – often – quality.

At some point, this potential community of suppliers and developers will run headlong into procurement, information security, requirements for assurance which hamper progress, government’s general plodding slowness, and all the rest of it. These are the fundamental problems that government has, and they’re the reason that all the people who currently work in government, on the web, have not produced anything like Alphagov. It’s not for want of trying. It’s not because they don’t know what they’re doing. They are a talented group of people, struggling against a system that’s more or less designed to stop them getting things done (particularly at the moment).

Those are the problems we need to solve. Making Alphagov was the easy bit. If it gets turned into the new HM Government website without solving those problems, the community won’t grow, change won’t be quick, cheap or easy, the innovation won’t happen. And after a few years, it will have turned from a bright new hope into another mediocre, monolithic lump.

That said: I am optimistic that this won’t turn out to be the case. As Mike Bracken has written, all the elements are lined up favourably. And I have a lot of faith in Mike, and Tom Loosemore, and Martha Lane Fox, and in the ministerial support they all have. But the biggest challenges are definitely ahead of us.