We Won: Global Datafest, Social Cycle, and Organising a 30-City Hackathon

There’s a particular kind of panic that hits at 2am when you’re running a hackathon and the WiFi goes down. It’s not “oh no” panic. It’s “forty developers are staring at me and I’m the person who said the WiFi would be fine” panic.

That was my Saturday night at the Global Urban Datafest, a hackathon I helped organise at FabLab London. And somehow, despite the WiFi, the cold pizza, and the fact that I was supposed to be an organiser, our team ended up winning the global prize. Let me explain.

What Was the Global Urban Datafest?

The Urban Datafest was coordinated between IBM, Stanford, and a network of urban data organisations across the world. The idea: run simultaneous hackathons in 30 cities on the same weekend, all focused on urban data. London, New York, Barcelona, Melbourne, Nairobi. The bones of 300 teams hacking on city problems in parallel.

My role was organising the London event at FabLab London in Bermondsey. This involved: booking the space (free, cheers FabLab), sourcing sponsors for food (£400 from local companies, plus IBM covered the coffee), getting datasets from TfL and the London Datastore prepped, and doing that thing where you pretend everything is under control while frantically emailing people who haven’t confirmed their attendance.

We had about 45 participants in London. Not the biggest of the 30 cities, but a solid turnout.

Organising vs Participating

Right, here’s a tangent but it’s important. There’s a massive difference between running a hackathon and being in one. When you’re a participant, you show up, drink free Red Bull, build something fun, and go home. When you’re organising, you’re dealing with:

• The projector that won’t connect to anything made after 2009
• The participant who’s allergic to everything in the pizza order
• The sponsor who wants to do a 20-minute presentation (absolutely not, you get 5 minutes, cheers)
• Power strips. You never have enough power strips. I bought 12 and still ran out.

I spent the first four hours of the hackathon not hacking at all, just making sure everyone had power, food, and functioning internet. The WiFi issue at 2am was caused by 45 laptops, 30 phones, and at least 8 Raspberry Pis all hammering the same consumer-grade router. We solved it by tethering to my phone. My data bill that month was not pretty.

Social Cycle

Around 6pm on Saturday, once things had settled down, I joined a team. I know, I know, organisers shouldn’t compete. But we were a person short, and I’d already set up all the datasets, and honestly I just wanted to build something.

The project was called Social Cycle. The team was me, Sarah Gallacher, and Vasilis Kostakos. The concept: map London’s cycling infrastructure against social deprivation data. Show where cycle hire stations, cycle lanes, and cycling investment existed, then overlay that with indices of deprivation. The question was simple: is cycling infrastructure going where it’s needed, or where it’s already affluent?

Spoiler: it was mostly going where it was already affluent. Quelle surprise.

We built a web app using Leaflet.js, pulling in Santander Cycles station data, TfL cycle lane data, and the Index of Multiple Deprivation. The visualisation showed a pretty stark correlation: Boris Bikes (as they were then) clustered in zones 1-2, cycle lanes concentrated in already well-connected areas, and the boroughs with the worst health outcomes had the least cycling infrastructure.

Winning

Each city picked a local winner, and then there was a global judging round. We won London, which was already a surprise. Then, a week later, we got the email: Social Cycle had won the global prize.

The prize was (and I’m not making this up) a certificate and a mention on the Stanford website. No money. No trophy. A certificate. But you know what, we were genuinely delighted. It validated the idea that simple data visualisation, done well, could tell a compelling story about inequality. You didn’t need machine learning or a blockchain. You needed two datasets and a map.

What I Actually Learned

Running a hackathon taught me more about event logistics than any course ever could. Some practical bits:

1. Budget for twice the power strips you think you need. I spent £48 at Argos and still had people daisy-chaining off each other.
2. Food is culture. Get the food right and people are happy. Get it wrong and they’ll remember nothing else.
3. Time-box the sponsor presentations. Five minutes, hard stop. People came to hack, not to watch slides.
4. Have a backup internet plan. Always.

The self-deprecating bit: I was so focused on organising that I forgot to take photos. The only picture I have from the entire weekend is a blurry shot of the pizza arriving. Of a global, 30-city hackathon that our team won, my photographic evidence is a Domino’s delivery driver looking confused.

If you ever get the chance to organise a hackathon, do it. Just buy more power strips than you think you need. And take photos.