Let it #uksnow, let it #uksnow …

I’ve always been just a little bit dubious and sceptical of these “grass roots”, “from the ground up”-type mashups. Do they really, genuinely happen? Can they ever achieve anything of real value, or do they just create toys for the geeks? Well, this week I saw such a mash-up evolving close up, and I admit – I’m a convert.

It started on Sunday afternoon, when the forecasters were issuing warnings of a “severe weather event” (honestly, is no corner of our daily lives free from corporate speak?) All we were seeing was a few isolated flakes, or the briefest of flurries instantly clearing up to be replaced by some glorious sunshine, which we shard with a few isolated tweets on Twitter. “Is this it?” we wondered; “Have the forecasters just been crying wolf once too often.

Then a few people in places like Cambridgeshire started reporting more sustained flurries; and the #uksnow hashtag started to appear, which I cheerfully adopted. Apparently I was one of the very early adopters, because I got a message from twitter trends aggregator twopular pronouncing me to be “among the top trend setter for trend ‘#uksnow'”. Oooh, the accolade.

And then up popped Paul Clarke who commented that it would be a good idea if these snow findings were recorded in some sort of usable format so that – if anyone fancies it – they could be mashed up with Google Maps to paint a real-time picture of the snow front’s progress.

I was sceptical that this would actually catch-on or get picked up by anyone, but I was game and started adding #uksnow KT6 1/10 to record a very light trace of snow in KT6 (Surbiton, South-West London) because it seemed like a nice idea to format posts if nothing else. And then whaddaya know, the #uksnow [Postcode] [scale out of 10] format was everywhere I looked.

The big test was whether anyone would actually take any of this spontaneously organising data and make something of it. Lo and behold, developer Ben Marsh did exactly that – creating the #uksnow Tweets mashup. And it was … very good. Really, very impressive for something thrown together on the spur of the moment. It even got a mention from the BBC on Rory Cellan-Jones’ dot.life blog. And what was really cool about this was that it had delivered a system that millions of pounds worth of space satellite was struggling to achieve: a real time picture of the progression of snow across the UK.

Since I procure web projects for a living, I couldn’t help but wonder how long a project like this would have taken to do commercially. Nothing fancy, just exactly this sort of project. I immediately envisaged hours of briefing and scoping meetings, discussions about functionality and design and interface. The procurement would easily have taken the better part of a month, and the project itself probably getting on for as long – even if anyone had found a workable solution. And yet instead, the whole thing had manifested itself – without project management, without a brief, without a client – and it had simply happened out the time and contributions of tens of thousands of people. It really was the most perfect jewel of an example of crowd-sourced mash-up potential being realised, enough to convince even the most hardened sceptic (as I had been) of its true benefits.

As a follow-up to this, on the Monday there was a Twitter challenge from Tom Watson MP, the Minister for Transformational Government, to reconfigure the Directgov site to carry news about school closures. He even bought a URL for the purpose, www.schoolclosures.org.uk, telling them that “if you did it for tomorrow am, you’d be heroes“. And could they do it? Well – yes they could, a feather in the cap for Directgov’s newly unveiled Innovate microsite.

Reactions to this have varied. Tom Watson himself was thrilled, Emma Mulqueeney put aside any quibbles to delight in the fact that it had simply been done, Simon Dickson gave a characteristically informative, interesting and even-handed account – and “Irish opportunist” Paul Walsh hated it.

The problem with it – compared with the #uksnow example – was that it just wasn’t organic. instead of evolving from a group of people, it had been the (admirable) work of a couple of developers in response to a Minister’s suggestion. There was no crowd-sourced information to rely on – all that could be done was to link back to the various council websites for information, so it came across as a rather unnecessary extra search engine layer which still resulted in frustration when the destination council site contained nothing more about what was going on than you already knew sitting at your keyboad 5 minutes ago.

The beauty of #uksnow was that it produced information where none was available before, and then used a Google Map mashup to deliver the results visually. But there was no new information for school closures, and it was the lack of on-the-ground information together with a centralised approach that meant School Closures lacked some of the “gee, wow” sense of the #uksnow adventure.

So why can’t we crowd-source information on school closures? Well, you can imagine how quickly that would be overrun and corrupted by hoards of school kids Twittering false information that their school was closed so that mum and dad aborted the school run. There was no such self-interest involved in mapping #uksnow, and any variations in reports quickly evened out statistically. But it’s a shame to have started a project like School Closures and then not think it through further – how to analyze the incoming tweets for frequency, reliability, contradictory reports, “trusted” accounts and the like to make it possible to gather this information from the ground and not rely on local council websites to get staff in to update their websites.

SchoolClosures.org.uk is a decent first step for Directgov’s innovate team, make no mistake – especially since it was just hours after their official unveiling. But let’s hope for some follow-through and sustained innovation on this and many other fronts, and not just quick “stunt” bursts that are quickly shelved and forgotten.


  1. paulclarke

    Excellent summary. A couple of thoughts – the first incarnation of the schoolclosures prototype was actually entirely reliant on organic crowdsourced data: it was an empty wardrobe with hangers taken from the Edubase data set, waiting for people to hang clothes on.
    What you’re referring to is a later service based on the local directgov “journey routeing” engine; providing a way for people to get straight to the most relevant deep link in their council’s website on the topic of school closures. A different paradigm, but one that could deliver some value by Thursday.
    What of the first version? Putting it out there provoked a torrent of highly valuable responses – everything from comments on functionality (relatively easy to add) to access controls (is there a way to stop kids closing their school – a little harder, but possible) to the data itself (very difficult, and possibly never wholly soluble, given the way that closure decisions are actually made in practice).
    In my view, the latter issue means that some sort of hybrid model of ‘official’ and ‘ground-view’ data is almost certainly the best route to explore.
    This development is still ongoing. It’s a little out of public sight at the moment; the snow came back, and people were starting to tell each other spontaneously that the prototype would help them for real now. Better to have some value from Thursday’s service than thousands of puzzled parents.
    But beyond the questions of performance of this service lie a whole load of deeper ones. As one example: how does the cheery, Blitz-spirit collaboration of volunteers, information enthusiasts and public servants relate to models of accountability, sustainability and commercial value? These are the most challenging of all.

  2. andrewlewin

    That’s really interesting about the first crowd-sourced version – I totally missed that one and never saw it. I suspect I was preoccupied with navigating South West Trains trying to get into the office.

    Great to have the inside track on it all – thanks Paul! And some really killed questions at the end, although frankly if I hear “Blitz” or “Dunkirk” mentioned in connection with this “snow event” ever again, I’m going to throw up … 😉

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: