The minimum requirements for blogging are a couch and a laptop; pajamas optional. The Gawker Media empire grew directly from these humble origins, but that was more than a decade ago, and now the company has a 15-year lease for three floors of a swanky Union Square office in Manhattan. When you visit, it’s strange to see “blogging” manifested in physical design. Every architectural choice caters to the spirit of sitting on your couch in your underwear with your laptop. So when Gawker’s new chief technology officer, Ian Fette, tried to find a room for us to talk in last week, he couldn’t, and we ended up sitting in a central lounge. (I was formerly an intern and then a contractor at the Gawker tech site Gizmodo, and I was definitely a proud member of the “no pants division” when I worked from home.)
Fette, who started at Gakwer in early August, is part of what company founder and CEO Nick Denton called a “second act” in a detailed memo about the company’s future at the end of July. “Gawker Media is growing up,” Denton wrote. Some of the changes he described, like the new office and hiring Fette, had been in the works for months, but others were motivated by the controversy over a gossip story that its flagship site Gawker published and then took down. “It’s been an interesting month, to say the least,” Fette says. But the 30-year-old Google vet didn’t really want to talk about drama. He wanted to talk about machine learning.
Gawker’s Kinja platform, custom-built internally for millions of dollars, has always had an identity problem. It simultaneously functions as a commenting system, a publishing platform, and a social network “with mixed results at best,” as Jonathan Mahler put it in the New York Times earlier this summer. In part, Kinja seems to be an attempt to mitigate the power social networks hold over publishers. Investigating Facebook’s influence in 2014, Max Read wrote on Gawker that, “If we’re not compelling our readers to ‘like’ our stories on Facebook, we’re missing out on traffic we’ve come to depend on.” If Kinja can be an ecosystem of its own, generating momentum and sharing from within, Gawker will have more stability, which could translate to a more consistent cash flow.
The ways Gawker could monetize Kinja have never been a secret—Ad Age noticed in 2013 that the terms of use specifically addressed where Gawker could place ads and affiliate links throughout the platform—but when Kinja was new, the specifics were still in flux. When Ad Age asked Denton about monetizing commenter pages back then, he said, “I find the topic irritating. It’s premature.”
Now, a few years later, it seems like it’s no longer too early to be having those conversations: It’s exactly what Fette was brought onboard to do. “There’s this big question of how do we get more readers to actually sign up with Gawker and get more engaged? And then once the users do sign up how do we better service those users? How do we give them a more personalized experience?” Fette said. “How do we do a better job of recommending to you what you should read next? How do we keep you on the site longer?”
Fette hopes his experience with machine learning will help answer these questions. He studied computer science at the University of Michigan and then did two years of graduate research at Carnegie Mellon, where he focused on machine learning and privacy. But then Google came knocking and Fette spent about eight years as a manager, first working with the Chrome group and then with Gmail—two Google products known for implementing predictive features. One project Fette worked on in grad school “was trying to use machine learning to detect phishing emails, which at the time—this was like 2005—was a very novel problem,” he said. “Now here I am at Gawker and we’re trying to figure out how do we highlight the best of the comments? How do we keep the spam out? So it’s very relevant.”
If Fette can successfully use machine learning to produce a better overall experience on Gawker sites and stronger reader engagement, the next goal would be nothing short of Kinja world domination. “We still have a ways to go before we can honestly say we are categorically the best platform for you to publish on,” he said. “But I think that’s where we’re driving towards.” It’s been a bumpy road. For example, Playboy, which notably published its Safe For Work vertical on Kinja for a year, stopped using the platform in 2014 in favor of keeping everything on Playboy.com. But Gawker has already invested in Kinja’s infrastructure, and if it can provide superior technology, it may be able to lure big names. As Digiday pointed out last year, in exchange for offering Kinja users access to the platform and a large audience, Gawker gets an “opportunity to populate its sites with stories from ‘premium’ publishers at no cost, and turn Kinja into a Facebook- or Twitter-like distribution platform unto itself.”
If Kinja is robust and user engagement is high, Gawker can not only support its own business, but can help other content providers monetize through Kinja as well. “The thing about having everyone on one platform is that we do have an opportunity to do more of this cross-promotion across the sites, and drive traffic, not just between the Gawker blogs, but potentially non-Gawker Media blogs as well,” Fette says. “How can we drive you more traffic?”
