How Zite’s News App Altered the Zeitgeist in Personalized Publishing
It’s just past noon on March 30, 2011. The entire executive team from Zite is at the Sharon Heights Starbucks in Menlo Park—the usual hangout for entrepreneurs making the venture circuit. They’re killing time between VC meetings on Sand Hill Road, and they’re a little nervous. But it’s not because of the meetings, or the coffee. It’s because of what they’ve just received in their e-mail: a cease-and-desist letter from a posse of 11 media giants—including Dow Jones, the Associated Press, Time Inc., and the Washington Post—whose content occasionally shows up in Zite’s personalized news reader app.
It’s only been three weeks since the launch of Zite’s iPad app, which has already acquired more than 125,000 users. But the C&D letter threatens to put a crimp in the Vancouver-based company’s growth, and in its fundraising efforts. The media companies charge that Zite’s app “damages our business by misappropriating our intellectual property” and demand that the startup stop using their content. Without articles from these organizations and their 300 newspaper and magazine properties, Zite would start to look thin indeed.
Half an hour ticks by. Just when it looks like things can’t get any worse, a leaked copy of the C&D shows up on AllThingsD, the Dow Jones-owned tech blog. Kara Swisher, the reporter with the scoop, calls the letter “legally lethal.” The team’s dismay level ratchets up a few bars. “We hadn’t even had time to digest it,” recalls Mark Johnson, an advisor to the company who would shortly become its CEO. “And we were like, ‘Sh-t, we are going to have to go into the next meeting'” with a well-publicized legal storm hanging overhead.
The team troops off to its meeting. But then something surprising happens: the C&D letter turns out to be a badge of honor. “The VCs were kind of chummy about it,” Johnson says. “They said, ‘If the big publishers are angry at you, I guess you guys have made it.'”
That soon turns out to be literally true. Over the next few months, the startup figures out how to mollify its media critics by changing the way the app displays their stories (it switches to a full Web view rather than a stripped-down reading view, sans ads). And just five months after the VC meetings, Zite gets snapped up by Atlanta-based CNN. These days, if anybody is still upset with Zite, they’ll have to take it up with CNN or its parent company, a little outfit called Time Warner.
Zite’s brief clash with the media establishment, followed by its swift assimilation into the same establishment, is emblematic of a larger story unfolding in the publishing business. Call it Aggregation Aggravation. Old-line media companies have spent the last 25 years coping with the digital fragmentation of their content, starting with CompuServe and AOL in the 1980s and exploding in recent years across the Web, RSS, Google News, Flickr, YouTube, Facebook, Twitter, Tumblr, and all the rest. Just as they’ve started to figure out how to make money in this radically altered world—not as much as they did in the print era, but something above zero— a new technology trend comes knocking: apps like Zite, Flipboard, and Pulse that hoover up publishers’ content fragments and put them back together in the form of tablet- and smartphone-based “magazines.” (See “The 10 Social News Apps You Need to Try,” 1/20/12.)
These apps typically use social or AI-driven curation to harvest a personalized mix of articles for each user. And they adhere to a minimalist design ethic under which the “extraneous” material on a publishers’ website—that is, advertising—is stripped away in the name of providing a more comfortable reading experience. Of course, the publishers’ initial, entirely predictable reaction was to cry copyright and trademark infringement and fire off the C&Ds. But now some publishers seem to be coming around to the aggregators’ point of view. And while the transition is a halting, muddled, incomplete one, it’s moving along much faster than previous media-business skirmishes (just compare it to RIAA vs Napster, or ABC, NBC, and CBS vs ReplayTV). “Frankly, as soon as we got those [cease-and-desist] letters we also got calls from publishers saying, ‘Hey, how do we work together,'” says Johnson. “At the end of the day, publishers realize that we offer something really unique. We are a discovery engine. We help people to find content they wouldn’t otherwise have read.”
Zite didn’t actually set out to disrupt the publishing industry. In 2005, it was spun off from the Laboratory for Computational Intelligence at the University of British Columbia in Vancouver as Worio—for “Web Of Research, Iteration One.” As that geeky name suggests, Worio offered a tool for enhancing Web searches. The software crawled and tagged Web pages and offered its own personalized page recommendations alongside standard results from Google or Yahoo. “The idea was that the relevance problem had been solved by the big search engines, but the interestingness problem had not,” says Johnson, who met with me in late February. “When you type Walmart, you are going to get Walmart.com, but not interesting documents about Walmart.” Worio found and highlighted those documents (assuming that they existed, which is a little hard to believe in the Walmart example).
But the startup wasn’t attracting many users or commercial partners. In late 2009, the company invited Johnson—a former product manager at technology companies like Sidestep, Kosmix, and Powerset who had landed at Microsoft through its acquisition of Powerset in 2008—to become a part-time advisor. “My assessment was that it was a really interesting research project without a focus,” Johnson says. “There was all this technology under the hood to allow great recommendations personalized to you…but it just wasn’t coalescing.”
At Johnson’s urging, the company refocused its tagging technology on news content. Recommendations, the startup reasoned, would be more valuable as part of “a personalized magazine where you’re looking for a stream of information in a certain topic area,” in Johnson’s words. The company renamed itself Zite and started testing an online news reader. But beta users still didn’t get it. “We were wringing our hands, wondering what to do with the Web product,” Johnson says.
Then, in early 2010, something big happened that provided the element Zite had always been missing: a context that would make it relevant to everyday users. “Lo and behold, the iPad comes out,” says Johnson.
Big publishers rushed to produce news apps for the tablet, while smaller developers brought out iPad versions of traditional Web RSS aggregators. These apps took advantage of the iPad’s reading-friendly design and its touch display, but they didn’t do much to help readers discover promising new content that wasn’t already in their feeds. “RSS readers are interesting, sure, but you have the same problem on the iPad that you had on the Web with Google Reader,” says Johnson. “We realized that the iPad was changing the way we read news, and that we had the technology to allow us to change the way everyone reads news. That was really exciting to us.”
The company abandoned its Web service and focused solely on the iPad. Johnson calls the app they came up with “a personalized iPad magazine that gets smarter as you read it.” (There’s now an iPhone version as well.) The first time you use Zite, you connect it to your Twitter, Google Reader, Delicious, and Read It Later accounts, so it knows what you’ve already been reading and what kinds of news your friends are tweeting about. You also pick a few “sections” from Zite’s predefined list of several thousand topics—Automotive, for example, or Luxury Lifestyle or Venture Capital. Zite packages up the most interesting recent articles its Web crawler has identified in your topics and assembles them on magazine-like pages.
Tap on a headline, and the app will take you to a decluttered reader view of the article, or, if the publisher has requested it, to the original Web version. The app pays attention to what you read and don’t read, and alongside every article are controls that let you give a thumbs-up or a thumbs-down, or ask for more articles from a given author or containing a given term.
Unlike its biggest competitor, Flipboard, Zite doesn’t rely much on social signals about what the people in your network are reading or sharing. It personalizes your sections by learning from your actual behavior. Says Johnson: “What makes us really unique is that we look at millions of pieces of content daily and from those, we try to find the hundreds of pieces of content that are most interesting in general, and then most interesting to you.”
When the free app finally hit the iTunes App Store in early March, 2011, the company wasn’t prepared for the magnitude of the response. More than 100,000 people downloaded the app in the first week. “After five years of having hardly any users and making lots of different tries at products, it was a real shock,” says Johnson. Zite’s engineers spent some sleepless nights keeping the company’s servers afloat. And within a few weeks—to circle back to the beginning of our story—they were getting invitations from Sand Hill Road.
Johnson left Microsoft in order to become Zite’s new CEO in late April, 2011, taking over from co-founder Ali Davar. Part of his job was to smooth publishers’ feathers. The problem that provoked the C&D letter was that in the first release of the app, “reader mode” was turned on by default. This got rid of all the clutter around Web articles, but deprived publishers of potential ad impressions in the process. “We did it because we have the bests interest of the users at heart,” Johnson says. Also, he says, there just wasn’t time to negotiate republishing agreements with every Web publisher whose content might show up in the app. “We are a small, scrappy company that didn’t spend a lot of time building relationships. We thought we would launch, and then talk to people. We had no idea we would get so big so fast and draw the attention of major publishers.”
The solution, though, turned out to be straightforward: if a publisher requests it, the app renders their pages in Web mode first. There’s still a button that lets users switch back to reader mode, but only after those crucial ad impressions have, at least nominally, been delivered. “There are some clear disadvantages to Web mode,” Johnson says. “It loads a lot slower. It’s not optimized for the iPad. User engagement isn’t as high,” as measured by how many times readers share or rate the content. But the compromise was the only way for the startup to get back to business.
And the more time publishers spend studying Zite, Johnson says, the better they understand how it differs from some of the other news-aggregation apps. Zite doesn’t have Flipboard-style publisher sections, for example, so it’s not as if the app is duplicating the experience readers would have at the publications’ own websites. Articles from one source are always interspersed with content from other sources—some of them quite obscure. “We look at hundreds of thousands of sources every day, mostly smaller publications you have never heard of,” Johnson says.
The topic-based content streams are likely to contain articles that wouldn’t turn up in a user’s existing RSS feeds or social networks. The way Johnson sees it, that means Zite is actually bringing publishers new readers. “We fare a lot better with publishers” these days, he says. “The e-mails we typically get now are ‘How do I get more exposure on Zite?,’ not ‘How do I change the way my pages render?'”
A few months after the iPad app debuted, Zite got a different kind of message, from CNN. “They said ‘We really love your product. How can we work with you?” Johnson recounts. “‘How can we work with you?’ quickly turned into ‘How can we buy you?'” Venture firms had offered Zite “interesting” term sheets, he says, but the CNN deal ultimately proved too enticing to resist.
The network promised to let Johnson run Zite independently and to leave the startup’s team untouched (today there are four employees working in Vancouver and nine in San Francisco). It said it would promote Zite on its website and in its TV programs, and that it hoped eventually to incorporate Zite’s personalization technology into other CNN apps. But it hasn’t leaned on the company for much technology help yet. “CNN bought us because they really saw something unique here, and very wisely, they are letting it run its course,” says Johnson.
With CNN and Time Warner as it new protectors, Zite isn’t likely to get many new C&D letters. That doesn’t mean publishers have gotten used to seeing so many people consuming their content via platforms they don’t control and can’t readily monetize. But that train left the station long ago—and where readers go, publishers must ultimately follow.
Johnson says Zite’s job is to keep improving the browsing and reading experience for users, while still helping publishers find a way to thrive. “I think any company in our position needs to be sensitive to the fact that there are three parties here—the application builder, the reader, and the publisher,” he says. “The application builder wants to make sure they are creating a great user experience and have all the right content and can deliver it to users. The users want to sit and read and have a really streamlined experience. And the publisher needs to get paid, at the end of the day. We have to figure out how to work together to make sure users are getting the content they want.”
In a world where readers apparently want to consume more and more content in bite-sized chunks on smaller devices, it’s not clear which new business models will ultimately work for publishers and app makers. “It might be subscriptions,” says Johnson. “It could be advertising, but it would have to be much better than current advertising on the Web.” Zite itself is experimenting with one form of advertising—Intel has signed on to sponsor the entire technology section of the app, and Bergdorf-Goodman sponsors the fashion section. “Or something might come along that makes micropayments a lot easier. It’s a hairy situation where no one really knows the end game.”
Aggregation aggravation, in other words, isn’t going away. But the publishers’ lawyers have—at least for now.