In every single case, the journalists and data analysts we consulted said there’s one crucial factor that gets journalists on board with data experiments, and it also happens to be the single most common trait in any newsroom: curiosity.
Experiments begin when someone looks at one data point and says “I don’t understand what’s happening there” or “that’s weird” and begins to form a hypothesis to explain the phenomenon.
But when there is a lot of data, it can be hard to spot anomalies. One good starting point: Ask newsroom teams to find a story or beat that’s doing better than it ought to do.
Social scientists call this “positive deviance,” “bright spotting,” or “outliers.” In newsrooms, it means identifying content that, without any additional resources, is outperforming similar content. Once you find an outlier, the next step is investigation.
It was a stats-loving sports editor at The Dallas Morning News who noticed that Southern Methodist University fans were acting a little weird. There are big college football teams in the area — the Longhorns, the Aggies. Texas A&M’s enrollment is close to 70,000, with a correspondingly huge alumni base. SMU, by comparison, enrolls about 16,000.
It usually takes about 3,000 pageviews for a regular visitor to the Morning News to convert to a subscription, a number that is much-discussed in the newsroom. Pageviews for stories about the relatively tiny SMU were lower than other schools, and the sports desk had begun to focus its coverage on schools that drove traffic.
“Our sports vertical editor at the time was our most metrics-driven person in maybe the entire newsroom,” said Amanda Wilkins, formerly at the Morning News and now a product manager for news tools at McClatchy. “He’s paying attention to everything around conversions. And what we discovered … was that SMU coverage from a pageviews standpoint was at the bottom of the list of colleges we covered, but people were converting on that content at a much higher rate.
“We weren’t doing a lot of coverage on them because pageviews were poor, but they decided to take someone off of the digital producer desk and turn him into an SMU sports writer,” Wilkins said. They asked the new sports writer to focus on one thing only: amping up content to get the SMU fans to the pageview threshold faster.
“It just worked,” Wilkins said. “It’s a small audience but the conversion rates continue to be a lot higher than the rest of colleges. … It worked and we started an SMU newsletter as part of that. It’s a really small audience, but open and click-through rates are over 50 percent for opens and 70 percent for click-throughs.”
So, although they frame it differently, journalists at the Dallas Morning News are aiming at the same targets as journalists at Financial Times in London and Whereby.us in Seattle and Miami: how much time readers are spending with content, and evidence that they’re forming a habit.
Short of mind reading or using a crystal ball to divine future prospects, most of the analysts and strategists we spoke to said that the most powerful predictor for how metrics adoption will go in a newsroom is whether reporters are interested in how their communities experience and respond to journalism.
Newsrooms that are curious about readers, and treat metrics as a sandbox rather than a report card, can keep moving closer to what they really want.
There’s just one more thing they need.