ProPublica’s Recovery Tracker, a database of stimulus funds broken down by state and county, makes it easy for anyone with the time and interest to browse through and find a story. Moving through pages and pages of data in any kind of comprehensive way, though, is a challenge.
Jeff Severns Guntzel, proprietor of MinnPost’s The Intelligencer blog, looked through the $7 billion of stimulus money sent to Minnesota and posted a short piece on an interesting aspect he found. “Although job creation was at the heart of this massive bill, there is almost no overlap between counties pulling in the most stimulus money and counties with the worst unemployment rates,” he wrote in a post on November 11. “Of the 20 counties that have pulled in the most recovery money, only one—Itasca—also ranks among the top 20 county unemployment rates.” At the bottom of the post, he asked for input from his readers about government contracts with an embedded Google Form tool.
But Guntzel knew that was just a start, and that there were more stories to be found. In his next post, he enlisted his readers’ help in systematically moving through the data, much like The Guardian famously did in tracking spending by members of Parliament. Guntzel’s headline asked, “Do you have 15 minutes to help us cover stimulus spending in Minnesota?” It was an experiment to see if he could get fifty readers to devote just a little slice of time to looking through the stimulus data and alert him to whatever stood out as something interesting to look into.
How much of a scoop can one find in so little time? Well, maybe not much, Guntzel admits. He says that the way ProPublica presents the information so clearly and searchably, it is very easy to get sucked in, looking for patterns. But “I’m assuming someone who would be willing to do fifteen minutes could get sucked in for an hour or so without realizing it,” he says.
To gather the information, Google Forms took the readers’ info and plugged it into a spreadsheet that was only visible to Guntzel. (It was a more direct and streamlined method than simply opening up a comments section, he says, which can get off track pretty quickly.) In exchange for their help, Guntzel would investigate whatever the reader had flagged, and the reader would credit in his next post.
Guntzel also says he hopes this process will help him put together a list of reader-volunteers he could potentially draw on for help with future stories on the topic. If someone is willing to join in the project for fifteen minutes, maybe they wouldn’t mind devoting a little more than that the next time around. “That to me is what is really exciting, bringing everyone in to the process. I would love to make the world’s longest byline whenever we write about the stimulus,” he says.
He also notes, though, that it has to be a truly collaborative experience for the readers to stay excited about volunteering their time. For that, he needs to stay in contact with them, ask them about the experience and what they’d recommend to other people who want to help out. “To be successful, this kind of thing has to go in both directions,” Guntzel says. “I can’t just absorb all the information and write a new post; I feel like I have to build a relationship.”
This same method has brought Guntzel success before; he’s only been at MinnPost for a few months (he was previously a senior editor at the Utne Reader), but he has already crowdsourced several projects. In one post, he shared a dataset of homicides in Minneapolis, and had his readers help sort the entries and map them in different ways.
He wrote another series of posts about Minnesota’s underfunded justice system; for those, he invited readers to share their experiences within the courts. “What you share will not be published, it will be confidential and will help to inform and guide our reporting,” he wrote. Guntzel says he was overwhelmed with responses about that, from people in all corners of the system: public defenders, judges, prosecutors, defenders, and court reporters. And those responses gave him a valuable overview of the important issues at play, shaping the focus of his reporting. He says it proved a far better method than the simple Google search that often kicks off a new story.
“When I’m reporting, I really like to just make a million phone calls, talk to as many people as possible,” Guntzel says. “And this is just such an efficient way to figure out the best people to talk to.” He adds the caveat that it’s no shortcut: “It’s an efficient way to get started, but I wouldn’t say it necessarily saves a lot of time, because I end up following up with everyone quite a bit anyhow.”
Each new project attracts a different subset of readers, so in a sense he’s starting from scratch each time. Readers who tell him about their experiences in the court system won’t necessarily be the same readers interested in combing through stimulus data. But each new project also expands his readership: justice department employees share the link with their colleagues at work, for instance.
Though the free technology that enables Guntzel to collect this information so easily may be revolutionary, the basic idea at play here—that readers can fill in the gaps of a story and contribute to a larger conversation with their feedback—is not.
“When I’m doing it, it just feels exactly like what I’ve always done as a reporter, which is listening to people, building relationships, and leveraging those relationships,” Guntzel says. “It doesn’t really feel any different from what I’ve done in the past.”
After our conversation, Guntzel followed up in an e-mail about “crowdsourcing,” the buzzwordiest of buzzwords, which he sees as just a 21st-century expression for a process that’s as old as publishing. He wrote:
Lauren Kirchner is a freelance writer covering digital security for CJR. Find her on Twitter at @lkirchnerI started reading this book “The American Language” by the delightfully disgruntled reporter/writer/editor H.L. Mencken. It’s the fourth edition. The first edition came out in 1919 and was 374 pages. The fourth edition came out in 1936, was completely rewritten, and was 769 pages.
What happened? He had read hundreds of letters from readers (one of them 10,000 words long) who offered corrections, critiques, and additions–and he listened. Those letters (and his own new discoveries) convinced him he needed to take another stab at it and guided him through the process. It was crowdsourcing.
