In August, Army CIO Lt. Gen. Jeffrey Sorenson will announce the winner of the Apps for the Army Challenge. He’ll be speaking about the progress of the contest at the Gov 2.0 Expo in Washington, D.C. next week. Regardless of which application wins the contest, however, the development of software by coders in the Army is an important case study in disruptive IT innovation. Creating application contests, especially for government entities, is no walk in the park, given the complex rules and regulations that govern procurement.
The man who may know the most about the nuts and bolts behind that process is Peter Corbett, the founder and CEO of iStrategyLabs an interactive agency based in DC. In the Government 2.0 world, Corbett is well-known for co-creating Apps for Democracy with Vivek Kundra, now federal CIO, and for his work on Transparency Camp, Government 2.0 Camp and civic entrepreneurship. An edited interview with Corbett on app contests, open data and innovation follows. For the raw take, you can also listen in to the embedded podcast.
What are your current projects?
Peter Corbett: At the moment, it’s Digital Capital Week, a specific innovation initiative we’ve got here in the District. It’s 3,000 plus people, all geeks and creatives, who want to make the city and world a better place. It’l be here June 11-20.
We’re also doing Apps for the Army. When application development challenges wrap up and all of the apps are shown, that’s when it really gets exciting.
Apps for the Army is important, in terms of what that might mean for contracting and for IT development in government. Where did the app contest phenomenon come from?
PC: We created Apps for Democracy with Vivek Kundra and the Office of the Chief Technology Officer back in 2008. Vivek had said, “Peter, I heard that you’re a guy that knows how to do ‘weird things’ with technology. I’d like to meet you and talk to you about what we’re doing in DC.” I came in and he showed me the data catalog, data.octo.dc.gov. At the time, I believe there were about 250 data feeds, all open government data about the District of Columbia, including real-time crime information, locations of potholes, school test scores, etc.
Two days later, I gave him a proposal and I called it, “Hack the District.” And he said, “No, that can’t fly. We’ve got to call it Apps for Democracy.” And three weeks later, we had a purchase order. Six days after that, we launched the contest. And within a month of launch we had 47 web, mobile, and iPhone applications developed.
That method spread. There’s Apps for Democracy contests in Finland and in Australia and Canada. There are ones on the city level in Portland and New York. And London is going to have one soon. There’s an Apps for Development by the World Bank coming out that is a global challenge. So it’s just a good idea at the right time that we intentionally made open. We told everyone how we were doing what we were doing so that they could do their own thing.
One statistic that’s been shared frequently is that by putting up X amount of money for a prize — whether it’s $5,000, $10,000, $50,000 — contest creators receive a multi-factor return on that investment. Is that accurate? What are the actual numbers here?
PC: D.C. government, after the first Apps for Democracy competition, estimated the value of the apps created to be in excess of $2.2 million. The way they made that calculation was by taking the amount of human resources time it would take to hire and procure developers to build all of these things. In addition to that, figure in the human resource costs of procurement managers and project managers staffed at the government, which they estimated to be a bigger piece of the puzzle.
That $2.2 million, is that figure real? It’s impossible to tell. Would the city have procured an iPhone application that shows you the nearest metro station and inbound and outgoing trains? They may never procure that.
Was it actually a cost-savings or was it a return on investment? It’s hard to have that be a very firm number because it’s not necessarily just about the applications developed. The actual engagement with citizens, and people becoming more passionate about doing things for their city, is almost invaluable.
Instead of thinking about new Facebook applications where you can throw sheep at one another, the technology community in D.C. is asking how it can build something that’s going to help their neighbors safely navigate the city. That cultural shift is invaluable. You have to try and assign return on investment figures to these things just because that’s the way it typically works in governments or large enterprises. But it goes beyond the hours and human resources required to do something that you can use.
Looking back at Apps for Democracy and Apps for America, how many of these applications are still around? How are they being used?
PC: We’ve done two Apps for Democracy competitions, in which there were 60-plus apps. I think only one of them has been turned off. That was a carpool mashup maker, where you could figure out better ways of getting a ride into the city. Most of them are still working and functioning. Some of them are used more than others.
One lesson I’ve learned is we don’t really know because we don’t have the analytics for each of the apps. If you’re running an application development challenge, it would be great to give your developers individual Google Analytics codes so you can track usage. We didn’t do that.
However, we do have information from the developers. DCHistoricTours.com, which is a way to build a walking tour of Washington, D.C., was incredibly popular during Obama’s inauguration. It was basically the only place that had a robust map of the inauguration route mashed up with Wikipedia entries and Yelp entries. They were seeing something like 3,000 or 4,000 uniques a day. If you extrapolate, that was more than 100,000 unique people using that application in a month.
I personally use Park It D.C. because I drive around the city all the time. It that helps you see where there are car thefts or broken parking meters. I use Are You Safe D.C.. That’s an iPhone application that shows you a threat meter, from green to red, on whether or not you’re in a dangerous place based on the number of homicides and robberies.
How is Apps for the Army different than Apps for Democracy?
PC: With Apps for the Army, there are 119 developers working on 98 applications. The largest bucket are Android apps. There are about 28 Android apps and about 26 iPhone apps being built . There are apps built on the open source LAMP stack. There are asp.net apps and Blackberry apps.
This is different from Apps for Democracy because of the nature of the Army. The compensation is employee cash awards for Army soldiers and civilians. We’re giving out $30,000 in employee cash awards, not prizes. The most crucial layer that we added in was the use of something called RACE (Rapid Access Computing Environment), which is a cloud computing platform that lets a developer build applications in a virtualized environment.
They’re building apps in the cloud in a secure space. What that means is we’re able to separate hardware from application development. If you know anything about federal, it’s so hard to get the hardware. If you’re a developer, how are you going to get an Android handset provision and then be able to do the work that you need to do? How do you get that to 100 people? How do you get that to people in Afghanistan and Iraq who are building stuff? We removed all of that concern by leveraging this cloud platform for development. There are certainly other idiosyncrasies, but that’s probably the biggest piece that’s different.
Most military software has historically been outsourced. With Apps for the Army, development is internal. If this contest is successful, what affect will that have on innovation in government and the military?
PC: You really have to credit Lt. General Sorenson for knowing and trusting that there was a latent demand for innovation among his soldiers and civilians. The gap between problem and solution is mostly being filled by contractors. That’s not to say there’s anything wrong with contractors, it just seems like there’s this default that’s been created over time where if you need technology you hire out.
But if a soldier in Afghanistan has a platform to build the solution to a problem through rapid means and have it certified and credentialed, they’re going to solve their problem so much faster. Arguably, they’ll solve it so much better because they’re the user. They’re the one with the problem.
What does that do for innovation in the long run? It speeds it up incredibly, which is what we need. We need to solve our problems faster and better than we are now because we have more and more of them every day. Rapid application development and agile development, which the startup world eats and breathes, is not only coming to federal, it’s there. We’re doing it. That’s very exciting.
If you were to step back from this and do an app contest all over again, what advice would you give to people or government agencies?
PC: I wrote a guide and told everybody how they can do it [available here and embedded below]. It was for selfish reasons because I couldn’t field all the phone calls and emails I was getting. The guide explains what I think a good apps contest or innovation challenge is all about.
[Below you'll find Corbett's presentation, "How to Create Your Own Apps for Democracy"]
Aside from that, what I’m preaching today is you’ve got to think beyond the apps. People get very fixated on applications. I get it. it’s been a successful model that spread so fast because you get tangible results.
What I’ve been talking about a lot — and I’ll be talking about this at Gov 2.0 Expo — is that the promise of all of this is to build civic innovation communities dedicated to solving the problems of citizens and government. Use the community’s talent and passion to turn open data into something different, whether it’s a web service or otherwise. And don’t forget that a civic innovation community isn’t just developers. It’s social entrepreneurs, designers, nonprofits, and it’s the government folks themselves.
We need to think about this from a long-term community development perspective. That’s in dramatic contrast to a short-term, 30- or 60-day application development contest. Those are important because they’re calls to action, but we can’t be too short-sighted. We need to think long-term.
D.C. has been in the thick of models for open data, open government and e-government, with efforts like the Open 311 API and DC.gov’s snow map mashup. What lessons can other communities learn from the civic-government ecosystem in D.C.?
PC: I think it’s great that cities are now competing with one another to be the most open and innovative. That’s awesome. That’s exactly what we should be doing.
Why D.C.? D.C. has been special because it has a leader like Vivek Kundra to put new things in place, such as project tracking dashboards or green-lighting Apps for Democracy. There’s also been a confluence of forces that were before my time. We have the CapStat program, which Baltimore pioneered more than 10 years ago. It treats city data like gold, putting it all in one place to make sure it can be correlated. What’s starting to happen is that people are building public data catalogs and more web services to make correlations between that information. D.C. has been ahead in that area.
I’ll be honest. There’s a lot of stuff going on in Europe that, if it comes to bear in the next three or four months, they’re going to leapfrog what we’re doing, which is cool. Now you’ll have continents and countries competing with one another to see who can be the most open and who can build the most interesting civic applications on top of their data. It’s great for the people. It saves money, and it accelerates innovative service development.
This interview was condensed and edited.
An earlier version of this post incorrectly stated that the winners of Apps for the Army would be announced this month. We regret the error.