On Monday night, Hillary Clinton and Donald Trump went head to head to make their case for why they should be America’s next leader. The stage also doubled as Facebook Live’s coronation.
The social giant partnered with ABC to live stream the debates via Facebook Live, putting the much-hyped feature on the national stage. Early reports suggested the debate drew a Super Bowl-sized audience. And Politico reported early Tuesday that 55 million people watched debate-related videos on the platform, building on Facebook’s success with the Democratic National Convention, when over 28 million people tuned into the live stream.
Outside of Snapchat, there is no more interesting corner of the social media universe than Facebook Live. Live has been touted as Mark Zuckerberg’s pet project, one he’s “obsessed” with. Some believe Live is the key to Facebook’s future—a resource that will help it compete against broadcast television. Others doubt that Live will ever take off. But no one can deny the potential of live video on a platform that has over 1.71 billion users.
What makes that potential even more incredible is that Facebook Live was almost shut down before the public ever got to see it.
The birth of Facebook Live
Facebook’s mythology is rooted in humble beginnings—a site started in the dorm room of a socially frustrated Harvard freshman. Similarly, Facebook Live began in relative obscurity.
After graduating from Columbia’s journalism school, Vadim Lavrusik joined Facebook in April 2011 as the journalism program manager. He launched Facebook’s partnership efforts with the journalism community and set to figure out how to make the platform more useful for reporters.
“I had been trying to convince people at Facebook to build live video four years ago when I was working with journalists,” Lavrusik said. “But the technology wasn’t quite there, and also I think people internally just didn’t believe it was going to be a consumer use case.”
But over the next few years, mobile technology accelerated. Data plans grew, and 4G LTE became widespread. In the fall of 2014, Lavrusik made his pitch again.
“We were able to convince our executives to give us one engineer,” he recalled. “I remember the fall of 2014, when we had recruited one engineer to work on it, [John Fremlin], and he had made some progress by December. We started to get more infrastructure engineers involved, and then Meerkat launched in March of 2015.” 1
People inside Facebook were still skeptical of the live-streaming project. According to Lavrusik, the biggest concern was whether Facebook could build something that could compete with established platforms like Periscope and Meerkat.
“My argument was always, look, we have the scale,” he said.
Lavrusik’s argument resonated. Facebook transferred him from the media partnerships team and made him a product manager for Facebook Mentions, an app that helps verified celebrities engage with their fans. Lavrusik and Fremlin pivoted the seven-person engineering team to work on Facebook Live, with plans for it to operate as part of the Mentions app.
“They were all working on different projects, and I was everyone’s least favorite person because I was getting them to work on something they didn’t want to work on,” he said. “Once we started making more and more progress, actually everyone got really excited.”
On August 5, 2015, Facebook launched Live on a limited basis to celebrities with a verified page. Even though the product “still wasn’t in a great place,” by Lavrusik’s own admission, it got its big break. The day after the launch, comedian Ricky Gervais went live from his bathtub.
It was a bizarre yet charming video capturing Gervais’s trademark awkward humor. Over 800,000 people watched it.
Right after the launch, Lavrusik showed the initial response to Facebook CCO Sheryl Sandberg and her direct reports. “Everyone had their ‘aha’ moment in the room,” Lavrusik said. “They were like, ‘Wow, we get it.'”
It didn’t take long for celebrities like Dwayne “The Rock” Johnson, Perez Hilton, and Carson Daly to broadcast using the app. Soon after, the team opened it up to all users, not just celebrities. And as they did, a treasure trove of data came back that would make Live even bigger.
“This thing is going to blow up”
Internally, Facebook monitors a metric that focuses on how friends share. The goal is to figure out how to get more users to post original content that’s native to Facebook, instead of just liking, commenting, or sharing a link from an external news outlet.
“Our U.S. release [of Facebook Live] basically moved that metric more than all of their other launches and efforts over the course of the previous year combined,” Lavrusik said. “At first I was like, ‘Oh, this can’t be accurate’, and then we had another analyst look at [the data] and it was real. I was like, ‘This thing is going to blow up.'”
Then Mark Zuckerberg saw the data.
Zuckerberg was already a fan of Live. But then he saw how much people were sharing live video compared to links and other video formats. He also noticed that, on average, people were watching live streams three times longer and commenting 10 times more than on regular video.
With Live, Facebook had an instantly valuable product. The videos were exclusive because users had to access the platform to create them. Since streams were live, the content was timely. And there was also an inherent interactivity to them, since people could comment as they watched.
“Mark was like, ‘Wait… why wouldn’t we make the video tab all about live video?'” Lavrusik recalled.
Live would give Facebook something it needed: a dedicated home for video. Zuckerberg planned to announce the new tab at Facebook’s F8 Summit the following April. Facebook’s product team liked the suggestion, telling Zuckerberg to expect a plan the following week.
Only days later, however, Zuckerberg emailed the team. He laid out product specs for the Facebook video tab and ordered the entire media engineering team to spend the next few months focusing on Live “in lockdown.”
Lavrusik’s team suddenly went from 12 people to more than 100. They worked through the end of March to make Facebook Live accessible to all users in time for the F8 Summit. Rumors spread that Zuckerberg was more “obsessed” with Facebook Live than he’d been with any Facebook feature.
Then, a couple of weeks before the announcement, Lavrusik left Facebook.
Since Zuckerberg announced Live with much fanfare in early April, telling users that it’s “like having a TV camera in your pocket,” the media has been considering one question: How big will Facebook Live really be?
Outside of all things Snapchat, Facebook Live has been the biggest social media story this year. After more than 800,000 people watched BuzzFeed explode a watermelon with rubber bands, Jonah Peretti, BuzzFeed’s founder, boasted that they had cracked the code, producing something online capable of topping the reach of live TV.
At the Digital Content NewFronts conference in May, media companies eagerly assured advertisers that Facebook Live was a game-changing platform.
To capitalize on that momentum, Facebook also paid approximately 140 publishers and influencers a combined $50 million to start creating video for Live.
But the content that might have shown the greatest potential for Live didn’t come from a media outlet or established publisher. Instead, it came from a 37-year-old mom from Texas named Candace Payne, or, as you probably know her, Chewbacca Mom.
On May 19, Payne live streamed herself in her car, unboxing a Chewbacca mask she’d just bought from Kohl’s. Her infectious laugh made it hard not to smile while watching, and Payne soon went viral as countless media outlets embedded the video on their sites. In total, it’s been viewed over 161 million times.
This type of success story highlights the potential of Facebook Live as a large-scale platform for both media organizations and professional content creators, as well as a place where amateurs and influencers can build an audience and go viral.
“The vision is basically, ‘How can we create the next generation of TV on mobile?'” Lavrusik said. “It’s going to start with this video tab because that will be a discovery surface for you to actually consume and discover interesting creators or content.”
Why, then, did Lavrusik leave Facebook while sitting atop the live-streaming world?
Facebook’s broadcast problem
“Most people think I’m crazy because they’re like, ‘Why would you leave your baby when your rocket ship is taking off?’ For me, I felt like that rocket ship was taking off, and I wanted to build another rocket ship.”
As The Wall Street Journal reported last month, Lavrusik left Facebook to launch a “one-to-few” live-streaming app called Alively. The startup was born, in part, out of user behavior trends Lavrusik noticed at Facebook.
When Facebook Live rolled out in the U.S., Lavrusik was surprised to see that teens used it at a much higher rate than anticipated. “We asked people who would try it once and wouldn’t use it again why they hadn’t used it again, and the number one answer that we got was along the lines of, ‘I want to be able share live, but I only want to share with a few people.'”
Facebook has known about this problem for a long time. The platform has become a place where you broadcast updates to everyone in your life, not share information only with close friends.
“I knew that we weren’t going to solve this problem when we made this strategic decision to basically punt on it and focus on making Facebook really good for broadcasting to lots of people—growing your audience, almost using Facebook Live to become famous, things like the Chewbacca lady,” Lavrusik said.
For now, Facebook has a legitimate challenge when it comes to the intimate sharing of live video with small groups of friends. That space seems destined to be dominated by Snapchat—which started dabbling with live streaming through its Live Stories feature this year—or by an upstart competitor like Alively.
The measurement problem
The biggest questions about Facebook Live come down to money.
Facebook clearly has a video measurement problem. Last Thursday, The Wall Street Journal revealed that Facebook had inflated the average viewing time for its video ads for more than two years, infuriating advertisers and damaging trust. When calculating the average time users spent watching videos, Facebook excluded sessions of less than three seconds. Facebook told Publicis Media it likely overestimated viewing time by 60 to 80 percent.
Even if the data was accurate, it would still be difficult to assess the value of live video for advertisers. BuzzFeed and Peretti, for instance, argued that the 800,000 concurrent viewers on their watermelon clip represented a turning point for online video, which could now compete with primetime TV. CNN, for instance, averages 723,000 primetime viewers.
The problem, as Kevin Draper explained on Gawker, is that Nielsen calculates television viewership on a per-minute basis. In other words, it’s an average of how many people are watching across each minute of the show. Draper writes:
Since it was broadcast live, the watermelon explosion has been watched by 10.7 million people, per Facebook’s count. If those people were as engaged as online World Cup viewers—and I’d venture that, on average, they were less engaged than people watching the most popular sporting event on the planet—those 10.7 million digital views would translate into an average-minute TV audience of 28,563 persons. If Peretti brought that number to advertisers at the NewsFronts as evidence of wild success, they would’ve laughed in his face.
Nonetheless, Facebook is forging ahead with a pricing model resembling television advertising’s. In August, Facebook rolled out the ability for publishers to insert 15-second mid-roll ads that would essentially act as commercial breaks for live streams. The feature is still in the testing phase, and Facebook doesn’t know if it’ll become permanent.
“That type of format doesn’t work as well for something like Facebook Live as it does for TV, because TV is completely a broadcast model where the audience can’t interact with the broadcaster,” Lavrusik said. “They can’t leave feedback.”
Lavrusik suggested that advertisers and Facebook could find a different solution, such as allowing advertisers to insert polls and other interactive content into mid-roll ads. However, the mid-roll format may not work as well for influencers and amateur video content creators, since they may not have the structure in place to sell and program 15-second spots. Facebook seems more intent on convincing those folks to use Facebook Live over other platforms like YouTube or YouNow, rather than setting up advertising infrastructure.
To tackle this problem, Facebook might adopt a tip-jar format, which has worked thus far on YouNow. Another possible avenue, Lavrusik said, calls for users to pay to have their comments and other interactions featured more prominently in the feed. He also said that branded content was a logical solution.
“What works for people who have huge audiences is different for these kind of digital influencers who might have smaller audiences,” Lavrusik said. “You need something that incentivizes up-and-comers to actually create content and grow their audience using this format. Then you also need something that works with existing models for big media companies and influencers with huge audiences.”
To virtual reality and beyond
As of today, Facebook’s hyped video tab has still only been released to a small percentage of iOS and Android users. Since Facebook has not released usage data for Facebook Live, it’s unclear how successful video has been on the platform. However, Lavrusik claims that before he left, Facebook had already eclipsed live-streaming competitors like Periscope, YouNow, and Twitch even before it rolled out Facebook Live.
“Facebook will win the one-to-many fight because of that scale.”
Despite the reach of its platform, Facebook’s biggest challenge for the future will be differentiating itself from those competitors, which is where virtual reality could come into play. (Facebook owns Oculus, which makes VR technology.)
“We started working pretty closely with our 360 team on Live 360, and I think that there’s going to be a big investment there,” Lavrusik said. “It’s going to take time for that to come to fruition. But you think about Facebook’s advantage, it’s the ability to move faster than some of the traditional media companies. They could figure out and jump into the live VR space before a lot of these traditional folks figure it out.”
Today, you might only opt to watch a presidential debate on Facebook Live out of necessity. But if Facebook Live were to become the dominant platform for live VR, that could be a game-changer. With 1.71 billion active users and growing, the platform’s potential is too big to bet against.
“Facebook will win the one-to-many fight because of that scale,” Lavrusik said. “Because you have so many people on your platform, you can deliver them the best audience and the most relevant audience.”
This piece has been updated to acknowledge John Fremlin, the engineer responsible for Facebook Live.