Analytics, giving the people what they want, and why that’s not always “clickbait”

What’s the difference between “clickbait” and serving the readers by giving them what they want?

Mathew Ingram of GigaOm asked that question last week after reading this fascinating piece at NiemanLab.org by ethnographer Angèle Christin.

Christin, after studying several newsrooms to try to learn how they’d been changed by the web, concluded that “all media sites now rely on web analytics to make editorial decisions. But this does not mean that they all use and interpret metrics in similar ways.” She also pointed out that journalists often say one thing about numbers and do another:

There is often a gap between what journalist say about metrics and what they do. Many writers express cynical views about traffic and say that they do not care about page views. Yet they almost always check whether they are in the “top ten” most read articles list.

Christin quotes a former New York Times writer named Richard Darnton, who says that in the ’60s, writers at the paper pretty much wrote for each other. Articles were assigned because editors wanted to see them. Little thought was given to the audience, other than to consider what they should want to read. Christin even points out that letters to the editor often went unread.

Compare that to what Will Leitch, then still at Deadspin, which he founded, said five years ago about why he didn’t apply for credentials at sporting events or write from the press box: “The minute I start doing that, I start writing for the other people in the press box.”

Was serving fellow writers and editors better than serving readers? The shorthand for the dangers of giving people what they want is Kim Kardashian, as in: If you give the people what they want, all you’ll ever do is write stories about Kim Kardashian. When I first started working online, in 1996, and we could first see exactly how many people were reading each story, the shorthand was Pam Anderson. Same point.

Yet somehow, two decades later, we have an awful lot of articles about Kim Kardashian, sure, but also quite a few that aren’t about Kim Kardashian. More than ever. As Ingram put it:

Has this transformation resulted in more clickbait and pandering? Undoubtedly it has. But it has also arguably resulted in more content that readers actually want to read, as opposed to producing reams of newspaper articles that no one ever makes it to the end of, just because some random editor thought it was important. And that’s probably a good thing.

“Clickbait” is a slippery term. Maybe we just all know it when we see it. But every use of analytics doesn’t lead to clickbait, and giving people what they want doesn’t have to either. Because one thing any set of analytics worth its salt will tell you the people want is: not clickbait.