American Journalism Review recently published a story about Ky Harlin, BuzzFeed’s director of data science, who created algorithms to predict which stories may go viral. While this piece hasn’t gone viral per se, it has received a high number of clicks by AJR’s standards.
At this point, though, almost anything with a BuzzFeed label draws a lot of traffic. One of the riddles of the Internet has been how to attract readers in a world bursting with information, and Harlin may have discovered the solution.
But BuzzFeed is, fundamentally, an aggregator and can pinpoint exactly what to include in stories to make them go viral. Traditional reporters and editors do decide what information is necessary to inform their readers, but they can’t change the facts or themes of their stories to better appeal to readers. Is it worth it then for non-BuzzFeed journalists to be so concerned about virality? My take: Good journalism should be the first priority.
Harlin’s not sharing his specific BuzzFeed formulas, but I had the opportunity to talk to journalists about a viral story that did not involve lists or GIFs: “Nightmare in Maryville,” published by the Kansas City Star in mid-October. The piece recounted the story of two teenage girls who filed sexual assault charges against two teenage boys and the uproar that followed in the town of Maryville, Mo.
It received about 2 million page views within 10 days, according to Kansas City Star publisher Mi-Ai Parrish.
This month, one of the accused boys, Matthew Barnett, pleaded guilty to endangering a minor.
Parrish said a big part of the initial story’s high traffic might be the pattern of sexual assault cases in the news. The fact that there have been other, similar recent situations — the rape case in Steubenville, Ohio, in which two high school football players were charged with raping an intoxicated teenage girl, the scandal at the U.S. Naval Academy in which lawyers “grilled a female midshipman about her sexual habits” after she had accused three male midshipmen of rape — helped to make the topic stand out in readers’ eyes, another Kansas City Star editor said.
“Nightmare in Maryville”‘s popularity could also be rooted in the details that made up the story.
“This one probably sold because of the sense of injustice, the youth of the victims and not least, the rural flavor of the crime…” Darryl Levings, a senior features editor at the Kansas City Star and the editor on the Maryville piece, wrote in an email.
Others said the quality of the reporting helped catapult the story onto computer screens.
“I felt like I could trust him [Dugan Arnett, the Maryville story author],” said Emily Bazelon, a Slate senior editor who wrote a blog post on the story. “He was speaking as someone who was being driven by his reporting, not by a sense of ideology, and it made the story very credible. I think that’s a lot of why the Internet writ large jumped on this story — it didn’t feel like vigilante justice run amok, it felt like really good journalism underlying a sense of justified outrage.”
In summary: What made “Nightmare in Maryville” go viral were the fundamental facts of the story combined with solid reporting. In other words, the same elements that have made certain stories popular for centuries — but can’t be manufactured out of thin air.
In contrast, Buzzfeed listicles can be constructed according to the author’s whim — or algorithm. It’s much easier to write a story that goes viral when you’re allowed to decide what 31 things will warm your heart on this cold day and include a picture of something described as “this teeny tiny adorable fluff.”
Arnett, the reporter who devoted seven months to the Maryville story, didn’t get to choose how much time 14-year-old Daisy Coleman spent passed out on her porch in 22-degree weather after the boy who police said assaulted her and left her there (about three hours). I doubt Arnett (who has not returned requests for comment on the Maryville story) included this fact to boost the likelihood that this story would go viral — though it’s one of the elements in the piece that could have contributed to such widespread public indignation.
Perhaps the problem is that I’m comparing apples to oranges, BuzzFeed to the Kansas City Star. But the two are becoming increasingly conflated as the definition of journalism evolves. As long as new media startups and establishment publications continue to co-exist, I worry that this obsession with ultra-shareable articles may overpower a key purpose of journalism: to inform readers about what’s going on around them.
Granted, BuzzFeed is starting to add more serious articles, and manipulating content is not the only way to produce a viral story. Harlin mentions that quantitative factors go into his algorithm as well, including “things like the amount of times something’s been shared on Facebook,” which clearly occurs after the piece is published.
I’d hope we could leave readers just as satisfied if we spent less time trying to figure out the secret behind sharing and more on producing quality journalism.
What if there is no pattern for producing real news stories that go viral? As Levings wrote in an email, “Folks are surely hoping to capture this wildfire in a bottle and light it again and again. We will see if the spontaneous can be harnessed. It may be Greekfire, for which the formula was lost.”
I guess this means we’d have pay attention, ask questions and listen to the people around us to get a sense of what we should be writing. That doesn’t sound so bad.