Apps Contest Winners Need Better Government Data to Sustain Innovative Services

Apps for Democracy and other contests generated by government open data programs need continued support.

by / May 2, 2011

Editor’s note: This story has been updated to correct some inaccuracies. The original version credited Daniel Odio with creating an application called Park It DC. That application which won an Indie Silver award from Washington, D.C.’s Apps for Democracy contest in 2008 was created by Shaun Farrell. The article also incorrectly said Park It DC had been discontinued. The application is still available.


A popular trend ignited in 2008 when the District of Columbia initiated Apps for Democracy, a contest challenging citizen programmers to use open data sets published by government agencies to make useful apps for the public.  

That year, the district offered $35,000 in cash prizes and got back an estimated $2.3 million in apps. Similar open data programs sprouted in New York; Philadelphia; San Francisco; Portland, Ore.; and others. Some of the programs operated as contests, and other cities just made the data available and promised to showcase the resulting apps. Applications came back to help citizens navigate public transit, find parking spots, locate government monuments — even to help map the safest routes for stumbling from bar to bar at night.

But now, three years after Apps for Democracy’s inception, some experts are questioning how useful the strategy has been.

New York City’s NYC BigApps generated a dazzling winner with NYC Way, an app enabling users to make restaurant reservations, buy movie tickets, book tours and more. The app’s three developers quit their Wall Street jobs to create MyCityWay, an application aimed at taking NYC Way’s functionality to cities across the country. The team has launched the project in the District of Columbia, San Francisco, Las Vegas and Los Angeles.

However, many apps generated by government open data programs cease being updated once hype over their creation subsides. Governments usually lack the interest and resources to take over maintenance of the applications, and developers have trouble making money off them.

Peter Corbett, CEO of iStrategyLabs, the district’s private-sector partner for Apps for Democracy, said open data activity is stalling in the United States. “You don’t see cities like New Orleans, St. Louis, Kansas City or any major metropolises jumping and saying, ‘We need to do this now.’ There doesn’t seem to be a stampede movement like there should be,” Corbett said, pointing out that open data initiatives are catching fire in European countries.

Former district CTO Bryan Sivak discontinued the Apps for Democracy contest after its second year due to the app sustainability concern. He agrees with Corbett’s assessment, but only to a point.

“[Open data] has sort of lost some of the luster it had a couple of years ago, but I think that’s typical for any sort of newer idea — new technology, new product or whatever,” Sivak said. “When you first launch something there is a peak of inflated expectation, which inevitably drops into the valley of disinterest.”

However, technology trends that follow that path often are later rejuvenated by innovative newcomers, he added. That could be the case with open data.

Several years have passed since the launch of Apps for Democracy, and it’s worth exploring how developers see the open data movement playing out in the real world. With many of the resulting apps going dormant after the initial hype, how should governments view the value of open data? And what do developers need from government executives to actualize that value?

Data that Would Make the Difference

Local governments may be posting enough data for developers to create novel applications, but not enough for them to make apps that a profitable number of citizens would pay to use, contends Daniel Odio, CEO and co-founder of mobile Web consulting firm PointAbout Inc.

Odio envisions dazzling possibilities for private-sector mash-ups of government open data, but doesn’t think enough open data is offered to create those mash-ups. Odio would like to see a world in which technology automatically delivers data a person might want before he or she asks for it. This already happens with applications like Internet radio site Pandora, which automatically sends an alert when a friend on Facebook likes the song we’re listening to.

Increased open government data will help spread that experience to other aspects of life, Odio predicts. He believes this will be realistic when mobile phones can connect to a car’s application programming interface (API). For example, with an API-to-mobile phone connection, as a person drives on a highway, his mobile phone app could detect that the vehicle’s gas tank is one-quarter full, which is the point at which this driver likes to refill it. Immediately the app determines the exit likeliest to please the driver.

“It knows you don’t want to buy gas from BP because of the oil spill, and so your phone tells you, ‘Hey, stop at the Exxon gas station that’s on the way to your destination,’” Odio said. “Literally the mobile device is letting you live your life in a better way. You can make better decisions because of this mash-up of data.”

Sivak agrees that developers need more data for better mash-ups, but he suggests that developers lobby governments for the specific data sets they want. He said developers need to give open data advocates within government good talking points to make cases for those requests.

“Part of the problem right now is we’re just saying, ‘Release data for data’s sake because we want to be transparent. Release data because somebody out there might do something with it.’ I don’t think that’s enough to really convince the bureaucrats that it’s worth spending limited resources on,” Sivak remarked.

He said a viable solution would be for governments to establish a formal process for submitting requests. Sivak cautioned, however, that developers would need to also foster relationships with like-minded individuals within the government who could advocate for releasing the data.  

Fixing the Format

Because local governments use different data formats, creating a national app is a laborious process, said Joachim Pfeiffer, creator of TransiCast, an app that’s fed by several local open data programs to help citizens navigate public transit systems. That, combined with the fact that numerous governments still don’t offer open data, makes it difficult for these apps to get a profitable amount of traffic, said Pfeiffer. “It’s purpose-built to transit, and it has a very high level of adoption,” he said.

To solve the inconsistent data format concern, Pfeiffer would like all transit agencies to use the General Transit Feed Specification standard.

Corbett agrees that a national municipal open data format standard is necessary for creating sustainable applications, plus it would make creating nationally appealing applications much easier. The absence of a national standard was among the reasons iStrategyLabs wasn’t interested in developing apps in the government open data market.         

Apps for Democracy winner Eric Gundersen views data format inconsistency as less of an obstacle. If agencies post data that’s valuable enough, developers will be willing to iron out the format hassles, he said.

“The fact that some of the data that is being released is dirty and not easily used is a market opportunity,” Gundersen said.  “We will go in and we will process data and package it up in a way that makes it easier to access, and then we can sell that.”

Speaking of “dirty” data, a common frustration among developers who use open data is sloppy accuracy. Pfeiffer said that while transit agencies are usually reliable about updating their open data and screening it for accuracy, that wasn’t always the case with other types of agencies. He thinks many local agencies aren’t using software to check data automatically before posting it. Pfeiffer suspected some of these agencies were using tedious, manual processes that were difficult to continue consistently.

“Symptoms are, for instance, that there is a boom and bust cycle. There is a big initiative to open the data, and the website is being created and the government is excited about it,” Pfeiffer said. “As soon as the attention is off, the feeds are not necessarily perpetuated in terms of being updated frequently.”

Pfeiffer declined to name any of these agencies. “I don’t want to point a finger at agencies that are actually trying to make a good effort,” he said.
 


Shaun Farrell

APP: Park It DC
CONTEST: Apps for Democracy 2008
STATUS: Partly updated, partly no longer updated

In 2008, District of Columbia-based Web developer Shaun Farrell took a coveted Indie Silver award in the District’s Apps for Democracy contest for Park It DC. The app informs motorists of what parking meters are charging at any given time. In addition to detailed meter pricing, the app generated data about crime against automobiles near the various spots and meters awaiting repair. Park It DC was initially a browser-based app and for much of 2009 it thrived on the Web, but due to popular demand Farrell created an iPhone version after the contest. In 2009, Farrell met with research and development staff at the District’s Office of the Chief Technology Officer (OCTO) to discuss ideas for enhancing Park It DC. Farrell wanted to add data alerting users about when and where temporary parking wasn’t allowed due to construction and when and where rush-hour parking rules kept them from parking in certain areas. OCTO staffers appeared interested at first, Farrell said, but that seemed to wane.  

“They seemed excited, but nothing ever kicked into gear,” he said.

In 2010, however, the district made drastic changes to the pricing structure for parking meters and expanded the time periods in which they charged. Some areas began charging until 10 p.m. instead of 4 p.m. Meters began charging on Saturdays, and prices rose in many locations. When Farrell attempted to upgrade Park It DC to reflect these changes, he found the data feeds had simply been removed on the government side.

“Everything was gone,” Farrell said. “There was no parking data, not even the initial feed I pulled from the data catalog.”

Farrell contacted OCTO to request a new data feed, but in late January 2010 when OCTO sent him updated meter data, it lacked the longitude and latitude coordinates necessary for plugging the data into Park It DC’s map. Farrell said further e-mail exchanges with OCTO didn’t produce much beyond that. He last communicated with OCTO in May 2010 when the agency expected another six to 12 months to pass before its contractor delivered the data feed. Parking meter data remains outdated on Park It DC, although the app still receives current data feeds on crimes against automobiles and meters awaiting repair.

For Apps for Democracy 2009, Farrell tried to follow up with his 2008 success with VacantDC, an app that displayed data on vacant properties in the district. The project didn’t place in the contest, but produced an unexpected benefit for some residents to contest property tax rates as a result of the district’s data feed sometimes falsely reporting their property as vacant. Properties in the district that go vacant for a certain period of time often have their property tax rates change. When occupants of those properties contacted Farrell to complain, he encouraged them to set the record straight directly with the city. The municipality’s mistake could have led to residents’ taxes being wrongfully changed, said Farrell. He no longer updates VacantDC, but it’s still on the Web.

These days, Farrell’s day job is as a Web development contractor for the Library of Congress. In the evenings, he works on Brewery DB, an online national database of beers and breweries. Farrell partnered on the project with two other developers in North Carolina. Brewery DB offers its data for free; its developers aim for users to offer their own data about beers and breweries. Farrell and his partners hope to use the data from Brewery DB for a suite of apps for the beer industry called PintLabs.  
 

Brian Sobel

APP: iLive.at
CONTEST: Apps for Democracy 2008
STATUS: No longer updated

In 2008, Brian Sobel and two colleagues, Travis Hurant and Tim Koelkebeck, won the District of Columbia Apps for Democracy contest’s coveted Indie Gold award with iLive.at. The app helps a person planning a move to the nation’s capital learn about nearby businesses, crime reports and demographics by simply entering an address. Before the contest, Sobel had been working as a software development contractor for the FBI. After four years he burnt out and quit, beginning two years of what he called an “entrepreneurial break.”

“About halfway through that break I happened to be in the right place at the right time when D.C.’s Apps for Democracy was a big movement,” Sobel said.

At a happy hour event, he met the two colleagues who helped him create iLive.at. After winning the Indie Gold, Sobel and his partners were turned down by the district’s Office of the Chief Technology Officer to maintain iLive.at for the city on a contract basis. Unable to monetize iLive.at, the developers ceased updating the application, and in April 2009, Sobel joined federal IT contractor Ntrepid Corp. Sobel said the potential of open data “Gov 2.0” pursuits felt played out for him after the contest, given the limited data made available by governments.

“It isn’t cutting edge anymore for me,” he said.

With only some cities offering open data, often in disparate data formats, Sobel said taking iLive.at national wasn’t practical. “I went back to federal government contracting because it was lucrative.”

He said he considered his time in the open data Gov 2.0 realm exciting and an experience that gave him tools and contacts that would likely be valuable in the future. Sobel foresees open data continuing to draw interest from fresh developers, but was cautious about whether it would prove to be a viable career shift for them.


Joachim Pfeiffer

APP: TransiCast
CONTEST: DataSF.org 2009
STATUS: Active

Joachim Pfeiffer still updates TransiCast, a mobile app he launched in 2009 to help users navigate public transit systems, with data posted on San

Francisco’s DataSF.org and other open data repositories around the country. Pfeiffer said he hasn’t been able to monetize the app because few local governments post transit data online and in a usable format to take the app nationwide. Pfeiffer continues updating TransiCast, however, because it benefits his job with consulting firm Booz Allen Hamilton, where he serves public transit agency clients.

“I use it as leverage in proposals to transit agencies that commission the development of mobile apps,” Pfeiffer said, explaining that TransiCast serves as an example of what’s possible for agencies.


Eric Gundersen

APP: Stumble Safely
CONTEST: Apps for Democracy 2008
STATUS: Discontinued

In 2008, Eric Gundersen and four colleagues placed high in Washington, D.C.’s Apps for Democracy contest with Stumble Safely, an app that used crime reports and bar locations to give pedestrians safe routes for bar hopping. At the time, Gundersen was running Development Seed, an open source software development firm targeted toward international development. It helped that Gundersen’s team was able to use aspects of a tool Development Seed was already developing for the U.S. Agency for International Development (USAID).

“We were able to leverage some of the same tools that we were simultaneously using for USAID to track bird flu, and we made a bar map with it. It showed how versatile some of these tools could be,” Gundersen said.

He took down the app shortly after the contest, but didn’t stop using open data. Development Seed helps government agencies refine and use their open and closed data to create maps for tracking elections and other humanitarian-motivated factors.

The firm is developing a map for the Department of Education that shows where schools reside in proximity to certain broadband speeds. The firm hosts numerous data sets taken from government open data programs, which programmers can use for free in their own Web projects. Development Seed recently announced a custom map-making Web tool called TileMill, which Gundersen hopes will be an open source alternative to ESRI’s Arc Server.

Andy Opsahl

Andy Opsahl is a former writer and features editor for Government Technology magazine.