Friday, April 24, 2009

Journalism's Online Future: What Made the 'Truth-O-Meter' Click

One of the best parts of sharing office space with the St. Petersburg Times' Washington team has been talking with bureau chief Bill Adair about PolitiFact.com -- the news site that he conceived and created at the start of last year's presidential campaign cycle, and that he is now nurturing into the first year of the new administration. Now that the Florida newspaper has become the first to collect a Pulitizer Prize for national reporting that appeared primarily on the Web, journalists should be carefully mapping Bill's DNA to try to figure out what his creation suggests about journalism's future.

Laboratory research will quickly reveal several genetic markers -- the evolutionary mutations Bill introduced that helped distinguish PolitiFact from a gazillion other political news sites, old and new. Among them:

  • An accessible, if not gimmicky interface;
  • A well-defined niche;
  • An important journalistic purpose and heritage;
  • And carefully documented reporting

Bill would be the first person to give credit to PolitiFact's many other contributors, particularly his newspaper's talented techies and designers, and a long list of reporters and researchers from the St. Pete newsroom and the Congressional Quarterly staff, all of whom can take pride in this week's award. Bill and the PolitiFact team also had enthusiastic support from the brass in St. Pete. (Disclosure: The Times is the corporate parent of my employers, GOVERNING and CQ Inc. As ever, the opinions here are mine alone.)

But this was Bill's vision. I remember when PolitiFact was just an idea that he crudely illustrated on the kind of three-panel cardboard backdrop commonly used to display middle school science experiments. The plan was ambitious -- perhaps even unsustainable, as I warned Bill two years ago, especially during a demanding and fast-moving election year.

But Bill was all but genetically engineered for this job. He had been writing about Washington for more than a decade, having already collected a prestigious Everett Dirksen Award for distinguished coverage of Congress. He also covered transportation issues for the Times and had written a book on airline accident ("The Mystery of Flight 427: Inside a Crash Investigation"). Writing that book involved organizing a voluminous amounts of fragmentary and often technical information. Bill kept the notes for his book in spreadsheets, using Excel's "sort" functions to sift and organize details and track sources as he needed them. That way of breaking down and reassembling information turns out to be a fundamental element of good non-linear journalism, in which the order or progression of information depends as much on reader choices than writer choices. ("More on this?" Click. "This or that?" Click.)

Did Bill metamorphose into a Webslinger by organizing large projects this way, or were his neurons and synapses just wired that way from the larval stage? Either way, he began building PolitiFact with the right editorial mindset. He did not start with an assumption that he needed to post scrolling pages of long-form text or beautifully shot and edited video or audio just to take what he was doing seriously. He set out to build a Web site.

This mindset was reflected in the underlying structure of PolitiFact. The site's core from the start was its Truth-O-Meter, which allowed users to sort its collected fact-checking by subject, speaker or ruling (true, mostly true, half-true, barely true, false and the dreaded "Pants on Fire!"). PolitiFact later added a "Flip-O-Meter" (with scores ranging from "no flip" to "full flop"). And the three-month-old "Obameter" was added just before January's inauguration to track roughly 500 specific campaign promises that Barack Obama made during his campaign. The promises are organized by subject and by status (from promises kept to promises broken, with other steps in between, including no action at all). And throughout the site, the PolitiFact team has meticulously listed and, whenever possible, linked to the sources on which they were basing their conclusions.

In short, PolitiFact provided dozens (I stopped counting at 192) of easy-to-understand and relatively easy-to-navigate ways for visitors to sort through the site's accumulated and carefully attributed findings -- along with an almost daily stream of more traditional articles that primarily serve as a way to connect dots or signal new findings in a small box atop the homepage. Without relying on user profiles, animated controls or other fancy forms of personalization, the site enabled its visitors to customize their experience, based on their interests and questions. ("Who's lying?" "How's my candidate doing?" "What about the others?" "And what's this debate really all about anyway?")

Plenty of other good sites take this approach to organizing content, particularly online services that provide news and information on finance, sports, weather, traffic, entertainment -- topics that perhaps more obviously lend themselves to the approach. But few apply this multilayer, user-directed model to other bread-and-butter journalism beats, such as local news, national news, foreign affairs and politics.

The data-driven political blog FiveThirtyEight was -- and is -- a good example of a more dashboard-like approach to organizing daily news. CQ Politics, PolitFact election-year sister's site, also did this to some degree with its continuously updated race ratings for the electoral college and all the 2008 congressional and gubernatorial contests -- a feature that was prominently displayed throughout the site, but that was still second to the day's this-just-in, headline-driven coverage.

The navigational approach/mindset I'm describing is much more common in project-based journalism. One of the most impressive examples is "13 Second in August," a Minneapolis Star Tribune special report presented with a scrolling aerial shot of the 35W bridge after its 2007 collapse. Clickable numbers on each smashed and abandoned vehicle take visitors to the stories of the occupants, using a combination of text and multimedia.

Could this approach work to covering breaking news -- say, for instance, day-of coverage of a disaster on this scale? Absolutely, as public radio station KPBS did using frequently updated interactive maps during the 2007 wild fires in San Diego. But pulling off such coverage means that, from the very start, editors and top producers need to think about how their reporting and information is organized -- not just how to get it or the order in which it was received.

I learned that at washingtonpost.com overseeing our overnight coverage of the presidential and vice presidential debates in 1996, 2000 and 2004, when we embedded a small, sometimes animated "Debate Referee" to serve as our fact-checker throughout each transcript. Clicking on the referee opened a window that had a short bit of text on the veracity of a candidate's claim and links to other articles and off-site resources that provided more information and explanation.

I warned Bill about the challenges we faced cranking out our quadrennial Debate Referee boxes, many of which benefited significantly from the traditional, long-form fact-checking articles produced each debate night by teams of reporters in the Post's print edition newsroom. Turning that kind of work into a day-in, day-out operation for the 18 months leading up to election day did not sound sustainable to me when Bill first started explaining his idea for PolitiFact. But he was undaunted and his enthusiasm was infectious. With his Science Project cardboard chart in tow, he convinced his editors in St. Pete to devote considerable resources to an editorial experiment they believed in.

The Truth Squad Tradition and Missing Links
One of the most shrewd decisions Bill and his editors made was to choose their niche. While driven in part by competitive reality, the choice to focus their online election coverage almost entirely on vetting the statements of the 2008 presidential candidates distinguished the site's content from the more ephemeral enterprise of chasing campaign polls, ground movements and other tactics. That focus also linked their site to a number of noble but generally under-appreciated editorial progenitors.

The tradition of "truth squading" took on greater urgency among political journalists two decades ago, after the fact-twisting TV ads of the 1988 White House race. That's when columnist David S. Border (my first boss at the Washington Post) began urging his colleagues in political journalism to "become more like consumer reporters," systematically scrutinizing the content of campaign commercials -- the dominant form of political communication at the time. As David put it in a 1991 speech, these "ad-watch" stories would help voters "decide what was true and what was false in the advertising, what was real and what was distorted."

The Post and many other news organizations, national and local, print and broadcast, joined in David's cause. One such notable journalist was Brooks Jackson, a longtime Washington reporter for the AP and Wall Street Journal, who created the template for on-air "ad-watch" and "fact-check" segments on CNN in 1992. Brooks later continued that work online on FactCheck.org, an ongoing project of the Annenberg Public Policy Center of the University of Pennsylvania. Bill has often tipped his hat to Brooks and his FactCheck'ers -- even as he was building upon and revolutionizing how this kind of journalism was organized and presented.

Bill also did not turn up his nose at the hard work of generating reader interest. He built up buzz and traffic for PolitiFact, tirelessly plugging its work in segments on CNN, MSNBC, Fox News, NPR and elsewhere. He often appeared with a geiger-counter-like "Truth-O-Meter" prop (pictured here) sitting in front of him. Bill also carefully monitored the Web site's traffic reports to make sure he understood how his visitors were finding and using the site. And he has worked closely with the Time's in-house search-engine-optimization guru in recent months to make sure his journalism is reaching users via the sites they use most. (How many news organizations have their own SEO expert on staff? That's a question Recovering Journalist Mark Potts has often mused about.)

PolitiFact certainly missed opportunities, too. The site provided prominent feedback links, but few ways for users to publicly engage the editors and each other -- a conscious and debatable decision, intended at least in part to help differentiate the site from becoming just another place for partisan name-calling and bickering. But being interactive means being available to interact.

The site also needs a business model -- something Bill and his editors are keenly aware of. I was very happy to see an ad on the site just a moment ago that was paid for by one of the major "advocacy" advertisers that help underwrite other politically oriented Web sites and publications. That's a start.

Can serious online journalism like this sustain itself financially? I believe it can. But for the business of providing journalism to evolve, its editors must as well. Journalists can no longer just do the equivalent of reading news copy into an open microphone. Depth and expertise have to become as important as immediacy. And we have to present our work in ways that make sense to the medium in which we're working -- not the media from which many of us came. Bill Adair and PolitiFact point us in the right direction. Now just click....

1 comment:

jessika.mk said...
This comment has been removed by a blog administrator.