An Article Has A Lifespan Of 37 Days, And Other Findings From Pocket

Nate Weiner, the CEO of Pocket--the "DVR of the web"--pulls back the curtain on his company's data to reveal surprising ways people now read stuff online.

Spend enough time talking to Nate Weiner, and you begin to realize he has a unique vantage point on web publishing. Whereas we might visit a given site like the New York Times and take a note of what’s most popular or shared, Weiner, the founder of Pocket, has a commanding view of what’s powerful and shared across the entire web. And not only that, Weiner--whose indispensable app allows web denizens to save articles they encounter during the day for later perusal--also has a sense of which articles people take an initial interest in, versus which ones they actually end up reading.

Around the Pocket offices, Weiner and others like to talk about “what they’ve learned from Breaking Bad.” Not to cook meth to make a better income--their successful startup should be taking care of that. But rather, Weiner thinks we’ve entered an era of web publishing that could stand to transform itself more into the freer marketplace of supply and demand that exists in a world of TV-when-you-want-it. You used to have to tune in Thursdays at 8, religiously, if you wanted to stay up on your favorite show. It’s a notion that’s already laughably alien to younger generations. It has also, arguably, enabled a marketplace of greater experimentation that opened the window for bolder dramas like AMC’s meth-dealing epic. Could something analogous happen for articles?

Nate Weiner

Already, the signs coming out of Pocket’s data---over a million and a half pieces are saved to Pocket each day--are intriguing, suggests Weiner. “One of the more fascinating things we’ve seen,” he says, “are the most saved domains at Pocket.” The data tells two stories: one, that legacy brands matter--the Times, The Wall Street Journal, and The Guardian still figure near the top--but second, that upstarts can have a major footprint. Someone might write a post on Medium that blows up. A tiny blog might post something to rival the virality of the work of a major brand. Certain “really loud curators,” like Maria Popova, the proprietor of Brain Pickings, can wield outsize influence.

The global pie chart of saves to Pocket are an interesting snapshot of reader tastes--but even those only show the surface. Just as interesting, suggests Weiner, is users' behavior after a story is saved. For how long do readers remain interested in that article after saving it? How much time can elapse before the article withers, forever unread, in their backlog of Pocket saves? Weiner decided to investigate the question with a recent, popular Brain Pickings story about the passage of time. Though the flurry of saving to Pocket lasted about a week for the story, the long tail on people reading it and sharing it was a span of 37 days. “For well over a month, that article was very active inside of Pocket. People were reading, sharing it with friends, and back onto Twitter,” Weiner says. Everyone will get their 15 minutes; the successful blogger is entitled to fully five weeks and two days.

This is the stuff Pocket CEO Nate Weiner is going to read later, using Pocket. Half the items were saved from Twitter, 40% were shared by his wife or a coworker, and the rest came via browser or email, according to Weiner.

The other interesting piece of information to come out of Pocket is the vast disparity between what is saved to Pocket in aggregate and what is actually read and appreciated once saved. The vast majority of items saved to Pocket are short-form stuff: blog posts, BuzzFeed listicles, quick-hit news items, links gleaned from Twitter, and so on. Eighty-seven percent of Pocket’s saves by volume are things of this nature. The remaining 13% constitutes longform stories--the weighty New Yorker piece, the thoughtful New York Review of Books essay, and so on. But while that 13% constitutes the minority of Pocket content, it’s arguably the soul of what goes on inside Pocket.

“Engagement around articles is heavily weighted towards longer form and higher quality content,” says Weiner. Metrics of “engagement” vary, but include: the likelihood of a user actually opening a piece, how often they open it, whether an item is shared, whether it’s favorited, and whether it was read through to completion or abandoned mid-scroll. Some articles become so vibrant in Pocket that they result in a share almost every time they’re opened; Weiner calls such stories “Pocket-viral.”

First publishers strained for the click; then the share; now, in the age of Pocket, they may need to try more holistic measures of “engagement” to gauge an article’s full worth. Pocket gives a more granular look within publications, too--for instance, two authors may be equally saved, but while one has an “open rate” of 10%, another may have an open rate of over 80%. “People are very loyal to certain writers,” says Weiner, whose own tastes run toward films, tech-industry news, and management advice. "You can see who are the authors who have the most loyal reader base, where people are hanging on their every word.” (Data like this is one reason Pocket launched a publisher program earlier this year to furnish publishers with basic data around their most saved content.)

Towards the end of our conversation, Weiner comes back to his favored metaphor for Pocket: that it’s a sort of DVR for the web, and that the lessons that apply to it are largely those that have been gleaned from shows like Breaking Bad. In the middle of a point about niche audiences, he says, “Think about the pitch of that story--a meth head who gets cancer,” and I have to ask--though Breaking Bad is reportedly a favorite around Pocket’s offices, is he personally a fan? “Breaking Bad I have not actually watched,” he confesses, citing a preference for House of Cards and Orange Is the New Black. “It’s funny. Everyone says it’s one of the best shows, so you know if you watch just one, it’s not an hour commitment, but 80 hours. You know you’re gonna get sucked in.”

Now that's behavior that will leave the number crunchers scratching their heads: in some cases, consumers may be most excited about the content they initially take the greatest care to avoid. And for that, there's Pocket.

[Image: Flickr user Ryan Hyde]

Add New Comment

6 Comments

  • Scott

    Yes this is an interesting article somewhat. What interests me is if this is a brilliantly placed piece of PR for Pocket to get the attention of Google for a later purchase. Just speculating. While data mining is interesting I agree with Rae that the idea of digital hoarding makes us less effective. Pending a buyout, I'm interested to see how much revenue this business will generate in the next 2-3 years.

  • Tom Foremski

    The test article is not a typical article so the results regarding sharing and lifespan don't apply to the vast majority of articles saved/read/not read.

  • Darren Burden

    Research from 2006 shows that 36 hours is the number for interest in a story on news sites, and it was a sample size of 250,000 ... http://physicsworld.com/cws/ar... - maybe sites like Pocket, Delicious etc increase the lifespan of these stories dramatically?

  • Rae Francoeur

    I've wondered, based on my own reading behavior, how often people actually go back and read what they've saved, and how much time they spend on saved articles. Saving articles is optimistic hoarding, since there's finite time in anyone's day for revisiting articles. There's always something new and interesting to discover every time you go online. I save what I think I'm going to need for work, though I do have high hopes that I am going to find time to read long features I do save.

  • Jason Hensel

    I've recently discovered Pocket's Text-to-Speech function, and now I listen on my work commute to all the articles I've been saving. It's taken away some of the guilt of hoarding articles to read later, and it goes by much faster.

  • Andrew Hanelly

    This is definitely an interesting piece, but I've got to groan a little because of the headline. And I only do that because I know this number will be misquoted for probably much longer than 37 days. The average article lifespan *in Pocket* may be 37 days, but in real life (on the rest of the web), articles can have a much, much longer shelf life. Take a look at this recent post that states 50% of Forbes.com web traffic comes from articles older than 30 days (http://www.journalism.co.uk/ne...

    Don't get me wrong, I think Nate is a smart guy and this is a fascinating POV. And I definitely understand you've got to put the lede and the hook in the headline - but I think it's a bit misleading to make a claim like this, especially in the era where 80% of people don't read past the headline.

    Anyway, i thought I'd weigh in with this caveat.