Nieman Foundation at Harvard
“Yeah, I think probably people shy away from 20,” BuzzFeed’s Jack Shepherd told me over the phone. “Twenty feels real weird.”
We were talking about lists, of course. Not listicles — Shepherd says, with some exceptions, listicles are not what BuzzFeed does: “A listicle, to me, is the lowest version of the art form. By lowest I don’t mean bad — sometimes they’re really, really great,” he says. “But a listicle, to me, is something that is literally an arbitrary grouping of things. Ten ghosts. Or 11 Songs We’re Listening to Right Now. Things where there’s no narrative that’s driving it.”
I had called Shepherd to ask him about list length — why some lists have certain numbers of items, and how they get that way. I was hoping there would be a science behind it, a big reveal that explained why some BuzzFeed lists get stretched out to 65 or 84, while others stop shy of 10.
Our curiosity was piqued by a side project that Knight-Mozilla fellow Noah Veltman calls a Listogram, a data analysis of the frequency of BuzzFeed lists by length — or, more precisely, quantity. Consciously or subsconsciously, BuzzFeed list authors tend to cluster their products around certain lengths, Veltman found:
Veltman said in an email that his curiosity was piqued when he started to notice an increasing number of irregular list lengths, and fewer round numbers.
“These odd lengths are sort of a new phenomenon. Most listicle makers still tend to stick to the top 5, top 10 format,” he wrote. “But when you see BuzzFeed listicles, they have these arbitrary lengths, and it always raises the question in my mind, why THAT number of items?”
If you read enough BuzzFeed lists, you’ll start to notice some of things Veltman noticed as well. For example, some lists don’t actually contain as many separate points as they say they do. They might have that many gifs or images, but there aren’t actually (for example) “34 Reasons Why Parent Trap Dennis Quaid Is The Hottest Movie DILF Ever.” That list includes:
1. He owns a winery.
5. He’s practically a cowboy.
6. HE’S A HOT COWBOY WHO OWNS HIS OWN WINERY.
The third of which is just an amplification of the previous two, if you want to get technical about it. (The hotness is made clear elsewhere in the list. This is serious business.) Veltman calls this phenomenon listflation. Via Twitter, BuzzFeed founder and CEO Jonah Peretti offered a counterpoint:
@veltman @BuzzFeedBen @brianabelson what about listdelation? e.g. This one is really way MORE than 13 -> http://t.co/jRC3kHMIft
— Jonah Peretti (@peretti) August 26, 2013
Anyway — so what did Veltman learn crunching the data?
Despite the perceived move away from round numbers, the four most repeated list lengths are 10, 15, 21, and 25. (One of these things is not like the others.) But overall, Veltman was surprised by how common it was to write lists over 30.
“I’d have expected lists longer than 15 or 20 to be much rarer, since it seems like a lot to sit through AND it seems like a lot more work for the author,” Veltman wrote. In fact, Shepherd mentioned to a period in 2010 when there was an internal, unofficial competition in the BuzzFeed offices to see who could write the post that exceeded 100 by the most items.
“I think what we found ultimately is the amount of time it takes to compile 200 of something could be better spent actually narrowing it down to 20 and figuring out how to make it a really rich narrative experience,” he says. “The megalist phase of 2010 was an interesting experiment, but one that has mercifully gone by the wayside.”
Brian Abelson, who collaborated with Veltman on the project and is also a Knight-Mozilla fellow, found a slight correlation between list length and how many tweets the list gets — the longer the list, the more tweets.
It’s probably not surprising that there’s as much superstition as there is science at BuzzFeed around what list lengths should be. In addition to an instinctual dislike of the number 20, Shepherd told me, “It’s long been a superstition in the business — for years — that an odd number will do better than an even number.” So it follows that Veltman also tried to find out whether certain authors invariably favor certain lengths, to see if they had any “pet theories” about what works well and what doesn’t. (Where well = viral, that is.)
“Certain editors get attached to certain numbers. I looked at my top ten viral posts of all time, and three of them have 21,” Shepherd says.
@veltman @BuzzFeedBen @brianabelson does the author bias account for the popularity bias? Confounding variables!!!
— Jonah Peretti (@peretti) August 26, 2013
Indeed, there are some common rules that are applied to list length. But first, it’s important to understand that there are different kinds of lists, which I tried to categorize from my conversation with Shepherd. “There’s such a massive variety of experiences you can have with something that is a headline that happens to have a number in front of it,” he said.
First, and most basic, is the listicle. “109 Cats in Sweaters is literally that,” says Shepherd. “There’s nothing more going on there.”
Second is the definitive list — the list that sets out to encompass all of something, like The 50 Cutest Things that Ever Happened. These lists have a strong tendency to go viral — 316,000 Facebook likes on that one so far — but they take a lot of effort.
“I didn’t just go and find 50 cute pictures that I liked. I thought back to my five years of scouring for cute animals on the Internet and the ones that really struck me and stuck with me,” he said. “Compiling that list took a huge amount of time and required having spent a lot of time looking at these things, whereas if you do something that is kind of more simple, you don’t want to put a big round number.”
And finally, there is the framework list — the list that only exists to structure a narrative. Take, for example, 54 Reasons You Should Go To A Dog Surfing Competition Before You Die.
The number 54, “in that case, that’s just a way of organizing this story that I’m telling about this amazing experience I had watching these dogs surf in San Diego,” Shepherd says. “If one of the items is that bulldogs are better surfers, and I have five great bulldog pictures that I’ll group under that number, it sort of doesn’t really matter.”
This is the most important kind of list for BuzzFeed, the one that helped it get past its reputation for creating Internet drivel and begin building a broader audience. This kind of list has the best content, but also the most arbitrary length — there are as many items as it takes for the story to be done.
“The thing that I’ve been saying recently about lists is a list is just a scaffolding for a story. It’s just a way of organizing information,” says Shepherd. “I mean, The Odyssey is 24 chapters. You could call that 24 Chapters About Odysseus. That’s, like, a really great list. Really top notch. Really, really viral. Super viral.”
That’s heady stuff. But the point is, list length has more to do with what’s in the list than with trying to guess which lengths readers are most likely to click.
Which doesn’t mean the lengths aren’t being manipulated. Lots of editorial thought goes into shortening or lengthening lists according to taste. “But it’s less interesting than numerology,” he says. “It’s more things like, ‘You know what, I looked at this post and I really felt like around 20, I got the point. So you might want to cut it from 31 to 21.'” (Not 20. Remember, 20 feels real weird.)
Lists being too long — or having too much mid-list dead weight — is definitely a concern. “We’ve looked at it in real time, which specific items on a list people were sharing the most, and started cutting out the ones that weren’t getting shares,” Shepherd says. “That was interesting — the idea being that, if there’s a dud in your list, you want to cut it down. But that didn’t in any conclusive way seem to effect the overall sharing of the post.”
Shepherd is also interested in using mobile technology to gather data on at which points in a list readers start scrolling faster. which could correspond with the point at which they got bored. He also likes to look over the shoulders of list readers.
“Something I would love to do — and it’s tough to get editors to do it when they’ve put a lot of time into a post — I would love to split test a bunch of different editorial posts. Let’s say an editor made a post that was 25 Steps to Get Through a Rough Day. I would like to have five versions of that post with different lengths and show readers different versions of that post and see which ones got the most sharing and see if we could get a sense,” he says.
But ultimately — much as one might want analytics to deliver conclusive answers — the results are often fuzzy. “I tried to have a look at some of our writers who have the most viral posts, and there’s actually a pretty wide distribution of numbers,” says Shepherd. People make lists of all different lengths for all different reasons. And why people click on them likely has very little to do with having a natural affinity for the number 12 or refusing to read lists longer than 30.
But thinking about list length does force us to consider the medium more seriously — something that BuzzFeed higher-ups have been trying to get us to do for a while. Ultimately, neither the magic of numbers nor science of numbers explains why people are drawn to lists to begin with — nor does length, on its surface, do much to telegraph what multitudes a list might contain.
“Honestly,” Shepherd told me, “I’ve often made posts where the post didn’t need a number, and then I’ll throw a number into the headline — just because people like that more.”
Photo by Thomas Hawk used via a Creative Commons license.
