I’ve been to a lot of hack days, hacks for the Houses of Parliament, for charity in general and for local government come to mind. I go because because it’s a way to help out a good cause, because it’s fun to try and get a working prototype of something done over a weekend, and because you meet interesting people.
Last Friday I went to another hackday and realised they don’t, in fact, help out the institutions that sponsor them. I know I’m not alone in feeling that this method of driving innovation is not working well as it might, and I know lots of other people are starting to wonder about it too. Of all the hackdays I’ve been to, not one has spawned a successful project. As far as I know, givey.com is the only example in the UK (I tweeted asking for more examples, apparently bufferapp.com almost came from a hack) This is an exceedingly low success rate.
I’m not a historian of the Hackday, but I believe the concept originates with a bunch geeky friends, friends of friends and colleagues bashing out code quickly – in essence techies that already know each other. In this scenario it works well. Certainly Google 20% time seems to be a productive version of this.
There is so much I could say about the hackdays – for example the problems when they become a part of a “corporate social responsibility” agenda rather than a way of getting things done. To keep this short, I’m going to focus on the misapplication of “Open Source” ideas and the obsession with Open Data.
Hackdays have been adopted by all kinds of institutions seeking digital innovation. These organisations are lost in the unfamiliar world of tech, as such they are empty vessels into which nerd ideology can be poured. Companies doing hackdays want to demonstrate how down with digital they are – often to the point of sycophancy. As a result Open Data and Open Source have become mantras whose precise meaning and relevance is lost – they become more important than actually getting something built.
Hackdays and the Open Source approach
In hack-land, the community driven nature of Open Source has translated into the idea that teams are liquid and that any attempt to define the project is anti-democratic. Exactly contrary to the Do One Thing Better Than Anyone Else, have-a-single-clear-offer approach that I take to be pretty much key to a successful project.
This conception of Open Source isn’t appropriate or relevant to hackdays. In any case, most open source projects seem to work by having a tiny kernel of people, perhaps just a single person, articulate a clear goal and demonstrate a significant ability to deliver that goal. Then a community of people who share that goal forms around it.
The “Open Source” mood at Hackdays means the reverse – you get a bunch of people who are not really connected at all and hope they can find a common objective between them. On more than one occasion I’ve worked into the night to deliver a working prototype only to demo it and retrigger an infinitely looping conversation about what the purpose of the project is.
Alternatively, in a follow up meeting a month after the hackday, someone will suggest something that fundamentally changes the project (for example: “Perhaps we shouldn’t be doing this as a website?”). The atmosphere will be the same as a brainstorm that should have finished after the first day: no idea is out of scope. In this situation it’s very hard to do any concrete work.
Magic Mashups – Open Data at hackdays
The tech world has evangelised about this for a long time and with good reason. There are important philosophical and moral reasons why data, especially gathered at public expense, should be publicly available, and also great reasons why it should be kept in non-proprietary formats.
However, if you go to a hackday this is not what you’ll hear about. At a hackday, you’ll hear that Open Data enables “mashups”, where combinations of disparate data sets unleash some world-shaping potential. The long-term architectural benefits of Open principles have morphed into a desire to have a data-based “mashup” in the App Store – whether it’s useful or not.
“We made primary school catchment areas and adventure playground coordinates available, what do you think you could build with it?”
The answer is, of course, an interactive map of primary school catchment areas and adventure playgrounds. Let use Open Street Map rather than Google Maps – high five! Usually the map will pique the curiosity and have little long-lasting value.
And Magic is in limited supply. If you have data about primary schools and adventure playgrounds then the resulting mashup is probably going to be about those two things. There is often an expectation that it will somehow be possible to draw out information that isn’t there – as if the median age of vulnerable immigrants can be coaxed from data on disabled parking spaces or similar.
The truth is that there are only a limited number of mashups that really bring something new. For the most part I think Open Data is going to have to be intermediated by journalists who dig out the stories in the data, it’s not a big public-facing thing.
Incentives – please, please be honest
The worst thing about doing hackdays is the feeling you sometimes get that the sponsoring organisation is doing it so someone can say “we did a hackday”. It just adds a ring of dynamism and sends a message that you’re down with all this digital stuff. A box successfully ticked – especially since there were no other ideas on how do digital innovation.
But when you aren’t paying the participants of a hackday the £600 a day you pay for agency work inevitably one’s eye wonders from the ball. Hackday briefs could often be so much better, and no time is allocated to evaluate the projects that come from the hackday.
The absolute worst thing though is the failure to say “No”. If you aren’t paying for someone’s time there is no reason to tell them that their project isn’t going anywhere. In fact, these people just did two days of free work for you, so turning round and saying “sorry, that’s not for us” seems rude.
Say: “Thanks guys, but we know our customers really well, and we don’t think they’d be into this”, or “Our board really needs to see something using the Open Data we’ve spent so much money on – your project doesn’t do that so can never be commissioned”. I might be disappointed, but now I can work on something else. Instead hackday projects are left to gradually loose energy and die a slow death, with no one having the balls to admit that the idea isn’t quite right.
When I first came across the idea of the hack I thought it was uplifting – people coming together to do work because they believed in it. I still think that, but that belief in the work is absolutely underpinned by the idea that it might go somewhere. It can never be a) Your startup, b) A project that does good in the world, then why bother?
Every hack is different, but I’ve been to a wide variety, and all the permutations of the model that I’ve experienced are failing to deliver.Google+