Last week, The New York Times announced that it was finally launching the first of its long-awaited news APIs.”Application programming interface” may sound complicated, but its essence is simple: the Gray Lady was packaging up its news in a way that can play nicely with others on the Internet.
APIs work by establishing a trusted relationship between computer programs so that they can share information, the way Google Maps are used by real estate Web sites to plot the latest listings. The first one launched by the Times, a presidential campaign finance data API, packages what the paper’s own reporters use to track the money chase of Barack Obama, John McCain, and third-party candidates and then sends that data out onto the Internet to see if it can make itself useful. The Times has announced plans to follow up with APIs of movie reviews, restaurant reviews, and congressional vote records.
Using Times newsroom resources to compile structured data might seem a bit off-track. But Aron Pilhofer, the paper’s editor of newsroom interactive technologies, puts APIs “squarely within our journalistic mission.” That mission is no longer to merely serve readers, but to serve as, as Pilhofer frames it, a valued peer in the Internet age. “We’re well past the time,” acknowledges Pilhofer, “when it’s enough to throw the day’s paper online.”
Like many papers, the Times has tried various digital strategies. Like many papers, it hasn’t been overly successful. Its attempt to port its old circulation model to the web, TimesSelect, collapsed in a heap of failure last summer. As it turned out, putting Tom Friedman and Maureen Dowd and Paul Krugman behind a firewall angered both readers—who were annoyed that even after paying fifty bucks they still couldn’t email links to their friends—and writers, who found themselves cut off from the party happening elsewhere of the Internet.
Then there’s TimesPeople, which is something like Facebook-lite. As an add-on to NYTimes.com, TimesPeople informs you which articles your friends think are the most interesting or newsworthy. As someone who devours the Times every day and regularly uses online social networks, I’m perhaps the poster child for TimesPeople. But I’ve logged in maybe five times, and made only one friend. I abandoned the service when that one friend’s lone story recommendation stayed stagnant at the top of my screen for weeks.
APIs, though, might represent the paper’s best hope of becoming an integral part of the modern news process—sometimes as a destination, but other times as a savvy information broker. As a trusted source for culled and polished fundraising data, for example, the Times can not only feed the stories of local campaign reporters across the country. It can also become a valuable partner in the growing online movement towards greater government transparency. Doing that might manage to keep The New York Times at the front of readers’ minds, and, just as important, keep eyeballs and links headed its way. Programmer Derek Gottfrid recently labeled the Times’s APIs “just another another syndication mechanism for us.” To make this work, it has to be more than mere syndication. The organization will have to decipher how to feed the Internet news that isn’t so easily packaged as reviews and finance data.
And they’ll have to do it while making sure they protect the quality of the news—which is what it will take to convince reporters to embrace the API revolution. In just a few years, we’ve seen blogs have gone from a quirky pastime for a handful of writers to an embraced way for journalists to track and tell stories. Similarly, some enterprising reporters will be attracted to and inspired by the chance to move away from the assumption that the cut-and-dried “article” is always the end product of the news process. With APIs, it becomes easy to imagine reported anecdotes piling up each other over time, until, perhaps, a narrative emerges. With APIs in play, savvy reporters, editors, photographers, designers and other content creators will learn to think about creating for them.
Think, for example, about what a “Mortgage API” launched by the Times two years ago could have meant. It would have given government reporters a place to feed those reported bits on Fannie Mae and Freddie Mac that didn’t fit into actual stories. Business reporters could have delivered curious data chunks showing something funny happening in the derivatives market. What’s more, enterprising readers could have introduced that feed to Google Maps API or Outside.in’s localization tools. Perhaps no one could have predicted the Wall Street meltdown, but in retrospect it’s clear that reporters had turned up useful information that ended up falling through the cracks.
Of course, it’s still unclear how to actually make money off of structured journalism. But the companies that have made money on the Internet—Google, Amazon, eBay and so on—have heartily embraced the idea that that they need to be enmeshed in the fabric of the Web. Amazon, for example, went out of its way early on to provide tools for online affiliates to market books and other products; the company’s S3 system, which rents storage space on its own cloud of Web servers, is now one of the Internet’s most popular data services. Meanwhile, Barnes and Noble started on the Web by replicating its stores online. The company has since scrambled to replicate Amazon’s approach to the Web, but according to Alexa Web rankings, barnesandnoble.com has about 6,700 inbound links while Amazon.com enjoys a quarter of a million.
“The Times has got to change,” Aron Pilhofer acknowledges with a laugh. “We have to start approaching the web as a medium,” he says, and not just a place to stick a Web site.
It’s an idea that’s inspiring in its counterintuitiveness, really. What might be The New York Times’s last best chance to thrive in the modern news business is to double-down on actually reporting the news—and then let the Web take it from there.