Modern data analysis shows old-school newspaper types were savvy about length

A post on Digiday, which covers digital media, marketing and advertising, caught my eye this week. The headline: Longer stories draw more attention, but with diminishing returns.

The diminishing returns have to do with advertising. If I’m reading it correctly, advertisers face a dilemma with longer pieces. A reader who scrolls down on a longer story is more likely to be engaging with the story, and therefore more likely to be spending more time with it, and thus more time looking at the ads. But think of all those readers who don’t scroll down, who just read a few paragraphs and then bail out. Advertisers aren’t crazy about placing ads where all those people will never see them.

But what interested me was the question of whether longer pieces drive more attention time, which, as writer Lucia Moses points out, “some are touting as the new metric of choice for digital publishers.”

Citing some analysis by the analytics company Chartbeat, Moses writes:

It’s tempting to think the Internet has not all but killed our ability to slow down and sustain our attention in an era of slide shows, listicles and other easily digestible posts. But the the truth is actually more complicated. It turns out that longer is better at drawing attention, but only to a point. Ironically, it turns out the ideal sweet spot for people’s Web attention span is about the length of a prototypical newspaper article.

While Chartbeat measures in pixels, Moses says that the ideal size Chartbeat found translates very roughly to about 700 to 800 words.

Maybe it’s just an accident that the old-school newspaper folks found their way to that length for a fairly typical newspaper article. But I wonder if their collective wisdom, something they’d come to understand about reading patterns down through the years, led them there.