It used to be that hackathons were organized by hackers, for hackers. Prizes, if there were any at all, were small. People played more for the glory and the challenge. Plenty of teams formed adhoc, more experienced hackers taking less experienced ones under their wing. A hackathon was a place where you brought that crazy hack you always wanted to try but just never had the time to experiment with. Instead of sitting by yourself on a Saturday, you sat down and worked on it with like minded people … over a couple of beers perhaps, sometimes late into the night.
This wasn’t a utopia that exists only in retrospect, my second hackathon ever our team stopped what we were doing for the better part of an hour to give an aspiring programmer, a woman of color, tips on books to read and languages to learn. As a clueless newbie hackathons helped me build relationships with technical mentors in what is sometimes a closed off, less-than-inclusive community.
A hackathon used to be a nice way to spend a weekend. Networking, socializing, and building … I used to look forward to them.
Now, of course, hackathons have become big business. At first the corporate component was unobtrusive – free food in return for a product demo – but more and more hackathons are being organized by people who have never actually hacked anything themselves. They pull the basic format, pump up the prize money, have legal add a whole bunch of rules to suit their agenda and expect the best and brightest to give up their weekend to work for free.
And as a result not only are hackathons not fun anymore, they’re also becoming rather exploitative.
Reclaiming the Hackathon
Way back in June we were asked to organize a hackathon for O’Reilly’s Data Week. For all the trouble that organizing a good hackathon can be, data hackathons are even worse. Understanding how to find and manipulate data are skills that rarely overlap with app building and rapid development. They should, but believe me they don’t.
So a data hackathon usually ends up being more of a data salon. A mix of activists, intellectuals, and data scientists talking about the state of the data. The lack of doing and building tends to alienate the hackers, and the messy, inaccessible state of open data confuses and frustrates them.
The challenge for us was: how do we organize a great event that will recapture the original feel of a hackathon yet not exclude data people who do not program? How do we get hackers excited about spending a Saturday working with open data when we know how awful the state of it will seem when compared to the tools they’re used to?
What we came up with was The Data Week Challenge, but what we learned from the experience is applicable to events of all kinds.
Sponsors Are Not The Audience
It can be tempting to organize your event around what will get you the most money from sponsors. After all, companies can and do turn a profit running these things, why shouldn’t you?
But consider why hackathons have evolved the way they have. Sponsors come for the hackers. Prize money is used to attract as many participants as possible. The larger the purse, the more expensive the event, the more sponsors needed. The more pandering to the sponsors the less attractive the event experience to the hackers, the more prize money needed.
Now consider the alternative: the hackers are the audience. If the event is built to attract hackers, then the sponsors will come and they will come on your terms.
Which is useful because some sponsors can be real pains in the ass, constantly telling you “if this part isn’t changed I will have to withdraw my sponsorship”. (Interestingly enough the amount of weight sponsors think they have to throw around is inversely proportional to the amount of money they are giving you). It’s nice to be able to say, “Go ahead, I can have your logo off the official site by the end of the afternoon”.
But What Do Hackers Want?
It was never a question for us what would get people excited about the event. Both Jacek and myself have been hacking in the community for years. But we frequently found ourselves having to explain and justify some decisions we made about what content we would or would not accept to other stakeholders in the event. There were a lot of hands involved in the logistics of The Data Week Challenge, and not all of them had ever been to a hackathon before, let alone knew the difference between a good one and a bad one.
If you happen to fit in this category, here’s a rough guide to things hackers like:
- Opportunity to show off technical skills
- Play around with new technology
- Hangout with other hackers, geek out over projects
- Build things just to see if they can be built
A good rule of thumb is: if you can hire a cheap Pakistani programmer from Elance to build it, it’s not technically interesting enough for hackers to want to give up a free weekend to work on it. People get paid to do what we do for a reason.
Hackers like to push boundaries, technical or otherwise. A project that requires you to outsmart built-in systems in order to force a piece of tech to do something it was never designed to do is fun. A project that is silly, mischievous, sarcastic, cynical, or quiet possibly illegal is great fun. Pointless is perfect as long as it’s funny, although useful can work too if it’s a passion project.
The hackathons that my friends look forward to inevitably break down into three categories: events that will offer them a huge platform to show off (TechCrunch Disrupt), events that allow them to hack on their passion projects (Music Hack Day), events that focus on testing their skills against others (college hackathons)
Everything comes down to the format. The format will determine what the value is to sponsors and who the sponsors should be. The format will determine what kind of space you need and for how long. Your hackathon format should be a reflection of what you hope to accomplish. Do you want apps built? Do you want to raise awareness? Do you want to break down silos?
Here are a few options:
- The Classic: Two-three day event. Demos of hacks built at the end. Judging, followed by prizes.
- The Battle Royale: Set of technical challenges, with increasing level of difficulty. Participants ranked based on ability to complete challenges, with optional bonus given for creativity.
- The Social: Often a minified version of the classic, one day only (generally eight hours or less), smaller group of participants. Post-event beers and mingling.
- The Worlds Colliding*: specific project areas advised by experts. Hackers and non-hackers work together to build potential solutions to real problems.
*This is fun when the experts and project areas are otherwise completely inaccessible to the hackers (rocket science! fashion! international spying!), less so when more pedestrian (education, nonprofits, publishing)
Sponsors: Who to Approach
So back to sponsors, how on Earth do you get a company to sponsor a hackathon? And how do you keep the sponsors from taking over?
The first problem is knowing who to talk to. Some companies like Github and 10gen have little sponsor request forms, this is the worst possible way to approach the issue. It takes whomever is supposed to be going through those submissions forever to get back to you, and only to set off another round of calls, emails and meetings.
The person you want to reach out to is the Developer Evangelist. Sometimes they have the authority over the developer outreach budget directly, but if they don’t they at least regularly speak to the person who does.
The nice thing about Developer Evangelists (speaking as a former one myself) is that we tend to travel in packs. We get sent to the same events all the time. So if I don’t know the Developer Evangelist for company XYZ, I can usually ask another evangelist for an intro. When I used to represent other startups, a lot of the events we sponsored came to our attention because other evangelists referred the organizers to me.
If the company is too small for a Developer Evangelist then you want to try to go as high up the food chain as you can. CEO is best, Head of Marketing will do … but Head of Marketing can be a tough battle because they’re usually given the least leeway with the budget of a young company. So if you have to convince the CMO, you also have to rely on how much influence the CMO really has.
As a side note: if you’re organizing a hackathon you’re probably doing it for the benefit of a company or organization. Trying to maximize the value of the event for your company, while still maintaining its ethicalness might seem like a problem, but it wasn’t in our case. We ended up using it as an ice breaker with people we wanted to develop business relationships with. Turns out people are a whole lot more receptive to conversations that start with “Are you participating in O’Reilly’s Data Week this year?” than ones that start “Can I tell you about my start up?”
Sponsors: What to Offer
Let me save everyone a lot of time here: No one cares if you put their logo on your T-shirt. Actually … make that, no one cares that there is a T-shirt. The sponsors that settle for just some branding can usually be broken down into two groups: companies where you’re personal friends with the decision maker and companies that are giving you something other than money.
Most sponsors prefer to do their sponsorship in trade (free space, food, content, etc) rather than cash, which is frustrating because obviously as organizer you’d prefer the flexibility of cash. But if you want the cash you have to offer the incentives, and the most desirable ones break down like this:
- Product Demo opportunity
- Recruitment opportunity (names, resumes, and emails of everyone in attendance)
- Pad user acquisition numbers for them
For the most part I think the idea of a super big company that just throws money at events is kind of an urban legend. Everybody wants something.
Sponsors: How Much Can You Ask For?
Short Answer: depends who you are. For an event that’s gotten some press coverage and some google-able buzz it’s not unreasonable to set levels in increments of thousands. But one of the most important things we learned is that the more unusual or untested your event is, the less you can ask for. It was perfectly fine for our friends running classic hackathons to ask for $1,000 ~ $2,000 per sponsor, but since we were doing something different in a niche not known for its hip and buzz worthy events the numbers needed to come down to the $500 ~ $700 range.
How Much Money Do You Need?
Less than you might think. Figure about $2,000 to $5,000 for the space (the cheapest price we got was $1,000 for a beautiful but out-of-the-way location, the most expensive was $40,000 for two days, including food. Around $1,500 to $3,000 for the food depending on how many meals.
But if you can get a sponsor to donate the space for free and another to donate the food, you can run an event on under a $1,000.
How Many People?
An idea that is getting thrown around among hackers in the community is to start charging a nominal fee for tickets to hackathons. First, because it eliminates at least some of the pressure to appeal to sponsors. Second, because it will decrease the no-show ratio. This was by far the most annoying part of running our event. The space really couldn’t hold more than 100 people, but we issued 150 tickets because we knew lots of the people would sign up and then forget or change their mind. Those sold out about two weeks before the event. The day of the event there were over 70 people on our waiting list.
A 70 person waiting list is kind of tempting when you know at least 40%~60% of the 150 won’t show up. If I had to do it all over again, I’d probably make tickets $5. A small enough fee that people won’t complain about it, but large enough to discourage ticket hoarders, wantrepreneurs and other undesirables.
Prizes, Prizes, Prizes
I have never come to a hackathon purely for the prizes, but I’ll admit it goes a long way to set the tone of an event. If you want geeks it stands to reason you should give away geeky things. The line between toy and tech is getting progressively thinner. It is super easy to find silly geeky things at Toys ‘R Us.
This is something we found especially useful in our case. Because we knew coming into organizing this hackathon that skill levels were going to be all over the map, we broke up the day with 15-minute workshops designed to help newbies with skills they would need to complete the challenges.
The big problem with organizing these was convincing sponsors that our rules about content were not cute suggestions. We were okay with product demos, as long as the product was a data tool hackers could use right away for free. We were not okay with talks focused on self-promotion, talks that had nothing to do with data, talks where the speaker was unable to answer practical howto questions. There were a few tense phone calls over this issue, but in the end it was worth it because we got what we wanted.
At the end the way we initially envisioned the event was an every-man-for-himself event, but most people preferred to work in teams, leveraging multiple and diverse skills to finish as many challenges as possible. While we didn’t envision this at the onset, it created a fun competitive dynamic that lasted throughout the day.
But that’s not to say we didn’t succeed in our goals. The feedback we received from the attendees was positive, a challenging day of fun, new technologies, and meeting new people, with good (if not too much) food in an all around relaxing atmosphere. A throwback to the early days of hackathons if you will.