Build Communities Not Contests

Part of the Open Government Directive provided instructions and encouragement for government agencies to hold contests. It even instructed the White House Office of Management and Budget to publish guidance on how agencies can develop innovation challenges in the federal space, and maintain compliance with existing procurement laws. As such, we’ve seen in just a few short months, a handful of innovation challenges pop up:

This list is off the top of my head and by no means exhaustive. From what I hear, there are a lot more out there and on the way. At Sunlight Labs we pioneered the development of contests on top of government data, starting with the “Sunlight Mashup Challenge” in 2007, then the rebranded Apps for America contests in 2009 and 2010.

Most government agencies are missing out on a core ingredient with their contests. Sunlight had a not-so-secret sauce in its contest strategy that ought to be shared.

For us at Sunlight, the not-so-big secret was that it was never about the apps. While it worked out great that there were apps like Filibusted and Know-Thy-Congressman, the production of the apps were never our desired end results. A win for us wasn’t the awesomeness that is, a great app built for Apps for America 2. It wasn’t even getting exposure to the technology community through the talks I got to give at places like Web2 Expo, and OSCon.

A win for our contest program was Jim Gilliam. Jim participated in our first apps contest with “Hello, Congress” that is now Jim entered that contest, became an avid member of the Sunlight Labs Google Group got inspired, and built not out of any apps incentive, but because he was inspired. Jim’s now building a business around that work.

See— the purpose of our contests weren’t to generate apps, but to build community. It’s to create a sustainable community of support and connection for the people who are eager to help out. The contests were an incentive model to build a long-term community of developers.

Fortunately for government, there obviously are already communities of developers starting to form around this data. The Sunlight Labs community, for instance is ripe and open for government participation. But in my days at Sunlight Labs, there was only one .gov address subscribed to the Sunlight Labs google group (though admittedly this could be because of agency policy, and more than one agency has participated in list discussions). There will be more. But just like there is no “The Government”, there won’t be a monolithic community of developers either.

Along with folks like the Sunlight Foundation and, Government should build developer ecosystems. Look to what Todd Park is doing over at HHS with the Community Health Data Initiative as an example. 18 months from now, as a result of keeping the eyes-on-the-community-prize, HHS will have a community rather than a bunch of apps. HHS will get its own Jim Gilliams — empowering new businesses to start, and ordinary citizens to take part in change.

If you build an apps contest for the apps, then you get apps. If you build an apps contest to build community then you get dividends like Jim, who’ll likely spend years helping to change the space. The apps that do get developed become more sustainable, and you’re building a platform for further, repeat engagement. Government ought to take this to heart and find ways to sustain the citizen developer’s interest. The point of the open government directive isn’t to litter the web full of disposable web apps that are soon forgotten about. It’s to build sustained developer interest around this data.

Get the book