YouTube's office is filled with its history. Inside the San Bruno, California, headquarters, about 45 minutes from the Googleplex in Mountain View, there's a YouTube video on every screen. Over here, the SmarterEveryDay guys talk about the brain-bending backwards bicycle. Over there, Rick Astley promises he's never gonna let you down. (YouTube's employees work in a semi-permanent state of RickRoll.) On a table in one of the office's many kitchens, there's a pile of remotes for Google TV devices underneath a handwritten “FREE” sign.
And of course, the red play button is everywhere you look: big doorways, small desk ornaments. The conference rooms are named after YouTube phenomena: Double Rainbow, It's a Trap, Dos Equis Guy, and on and on and on.
Inside the Lolcats conference room, VP of product management Matthew Glotzbach is describing the future of YouTube. He envisions an app so good, an algorithm so perfect, that it knows exactly what you want to watch at any given time. You wake up in the morning and catch up on the news while you get ready. Then, throughout the day, YouTube shows you shorter videos when you're waiting in line or in the bathroom: maybe some gadget reviews, or the best Jimmy Fallon bit you missed last night. At night, you come home, and use Chromecast to watch a movie or an episode of Video Game High School on your TV. YouTube wants to be more than a search engine for video. It wants to be the future, a perfect blend of TV and the internet, where everything is on demand but there's always something on.
A decade after its debut, YouTube is a behemoth. It's become the place for video online. Three hundred hours of video are uploaded every minute, and it has well over a billion users worldwide. It's spawned a crop of celebrities, real honest-to-goodness famous people. It's by some measures the world's second-largest search engine. And it has pioneered entirely new ways of creating and consuming video. Video was ascendant in the last decade, and it's going to be inescapable in the next one.
YouTube can't relax, though. Not yet, not ever. New challengers---everyone from Facebook and Snapchat to Vimeo and Vessel---are eyeing its talent and ready to poach its viewers. Absolutely everyone is coming for its advertisers, who have untold billions to spend and serious demands about where it goes. YouTube needs to prove it can turn impossibly huge view counts into actual, real profit.
The plan? Make sure everyone on the planet can get online, and on YouTube. They're working with carriers and ISPs to figure out how to stream to anyone no matter what their connection looks like. Then, get so good at showing them videos they like that they'll never want to turn off. That requires teaching their computers what's inside your videos, what videos you want to see, and what formats and video types are coming next. The video industry moves fast, and YouTube has to stay faster.
Simple, right?
Buffering is the dirtiest word at YouTube. The people who work there say it a lot, always with a sort of cringing, pained look. It's like they're remembering a bad breakup or just woke up to a crushing hangover.
"There was some stat that we used to have that was like, if the YouTube buffer symbol was a webpage, it would be the third most popular website in the world," Glotzbach says. Then he hastily adds: "Not now, though!"
That's been the mission since even before Google bought YouTube. You've probably seen the first video, from January 2005, of "me at the zoo." This is YouTube's "just setting up my twttr" moment, and has become a key part of the mythology. You probably also remember the video a year later, when Chad Hurley and Steve Chen announced they'd sold to Google for $1.65 billion. YouTube grew furiously, giving people an easy way to upload and share video that previously would've resided only on Handycams and Mini-DVs, never to be seen by anyone. It wasn't complicated or powerful---that was the whole point.
With Google, YouTube hit the big time. The acquisition was an obvious one; the two companies desperately needed each other. "Google Books was going on at the time, ingesting all the books," says YouTube's VP of engineering John Harding, who was one of the first Google Video employees. "And we had said, what would it take to have all the video?" Google built the tech for Google Video, scaled it to infinity, and had absolutely no idea what to do with it. Meanwhile, Harding remembers, "YouTube had a very fantastic user product." It made it easy for people to upload their own video, and its popularity exploded from the get-go. But YouTube didn't know how to scale; it was buckling under its own success. It was a perfect marriage, even if Google Video did live on in awkward redundancy for a few more years.
https://www.youtube.com/watch?v=QCVxQ_3Ejkg
Harding switched from Google Video to YouTube a few weeks after the acquisition. (After the YouTube acquisition, he says, "nobody was that interested in working with Google Video.") When he started talking to the YouTube team, he heard they were trying to figure out how to satisfy people who wanted to watch videos on their new Video iPods, PSPs, and even their TVs. It was exactly what Harding had been thinking about for the last year. "I was like, 'eh, I’ll just come over here.'"
It became the project that would dominate his career: how to get YouTube to as many people, and as many devices, as possible. In early 2007, the iPhone and Apple TV became the first big upgrades, forcing Harding and his team to convert every single video ever uploaded to YouTube to a newer codec and higher resolutions. "It seemed like a big deal at the time, but I don't know…" he says, trailing off. It turned out to be comparatively easy: He'd have to do the same thing many more times, only with unimaginably larger amounts of video.
The upgrades never stop. They can't, not if YouTube wants to stay ahead in a world where tech and money move faster than money at a Vegas table. YouTube has built new codecs, participated in developing new versions of HTML, and worked even neighborhood-by-neighborhood with internet providers to make sure people are getting the best picture possible. In 2007, YouTube upgraded to display 360p videos that would look good on the iPhone. Soon it was 480p, then 720p. Then Harding just said screw it. "In 2010, we announced 4K support, but we actually supported up to 8K at that time. And I mean, at the time 4K displays weren’t even a dream."
After years of playing a never-ending game of catchup, YouTube collectively decided to stop rushing to support the newest tech. They jumped way ahead of the field instead, and hoped it might even inspire others to catch up. In 2010, a YouTube engineer had to rig up four 4K TVs in a two-by-two square just to see if the newly-supported 8K footage would play. By CES 2015, 8K TVs were everywhere.
Now of course, Harding is turning his attention to 360-degree video and its implications in virtual reality and beyond. (He demurs when asked if there are VR headsets all over the office, grinning before offering a pat "Nothing to announce.") He's thinking about high frame rates, and HDR video. Glotzbach is excited to tell me that YouTube is about to start supporting livestreaming of video games at 60 frames per second, which is going to be huge. And they're both thinking about how to get all these videos in front of as many people as possible.
None of that matters, of course, if nothing happens when you hit play. Upwards of 80 percent of YouTube's viewership is outside the US, and in many of those places, bandwidth is both expensive and unreliable. "You go to a country like India," Harding says, "and sometimes you’re going to have 3G, but sometimes you’re back to the early 2000s of 2G." So YouTube started working on what's known as "adaptive bitrate streaming." It means your video can change its resolution on the fly, so it'll keep playing even if there's a network hiccup. It's taken a decade to get the data and support it needs from browser makers, smartphone manufacturers, and even ISPs, but what YouTube called "Project Sliced Bread" is now in practice. "It cut buffering in half in some places," Harding says. They're also working with governments and carriers to try and bring down the price of cellular data, and making videos available offline where coverage is bad.
That's part one: Get people online, and make sure there's a version of YouTube for everyone. There's another side of the plan, though, and it's harder than you think in a near-infinite sea of content: Give all those people something to watch. And then something else, and then something else.
Over the last few years, YouTube has shifted its focus away from single, viral hits---the David after Dentists, the Keyboard Cats---and onto the long-standing set of videomakers it called "creators." Glotzbach is broadly in charge of YouTube's Creators program, which manages and cares for the people who actually make videos. He's like an encyclopedia of YouTubers: In a single hour, he references conversations with everyone you've ever heard of, including the Gregory Brothers, John Green, Epic Rap Battles, and a guy named Sammy from Germany who's apparently become a gigantic YouTube star. And, of course the paragon of YouTube stardom, PewDiePie, the gamer-comedian whose subscriber count is pointless to print because it grows so fast. (I just checked: 36,489,262.)
The goal is to empower those creators, and then to sell them as celebrities---and charge advertising premiums to match. But that doesn't tell the whole story, says Brian Wiester, a senior analyst at research firm Pivotal. "If they want to capture TV dollars, they need TV content," he says. "That shouldn’t be rocket science." Until now, he says, YouTube's been so large and so entrenched that it's been able to focus its effort on monetizing its existing, lower-cost content. They've built better tools for advertisers, better user-tracking, and developed MCNs designed to act like record labels for YouTubers. And there's some money in that, sure. But YouTube has never proven it can sell the ads, or accumulate the right viewership, to support truly big-budget entertainment. "When I say premium," Wieser says, "I mean NFL football. Where is that NFL contract, anyway?" That's where the big money will come from---and while YouTube has proven it can turn a one- or two-man show into a real business, it can't take down TV until it looks and feels like TV.
In some countries, YouTube is already the place for re-runs of TV shows, and you can expect YouTube to make some higher-budget bets soon enough. "They do have an interest in developing a successful commercial property," Wieser says. "That seems to be clear." Glotzbach acknowledges this, but insists it's the YouTube-native community that will always be the stars of YouTube, the faces viewers and advertisers alike come for. They're also YouTube's best answer to competition from Facebook, Vessel, Snapchat, Twitter, Periscope, and every other company with a desire to make money. Of course, most YouTubers are on those platforms, too---social media is a double-edged sword for YouTube. On the one hand, they're great promotional tools, bringing people back to videos. "YouTube is still that nucleus, or that center," Glotzbach says. "A lot of our creators use other platforms as almost marketing, to bring fans back." But on the other, what happens now that Facebook has decided that rather than sending people to YouTube, it wants to host videos itself?
Glotzbach hopes the answer is as simple as this: YouTube pays its video creators. Facebook doesn't. But he and his team are also starting to think about how to get people to post their goofy behind-the-scenes videos to YouTube instead of Instagram, and even how to extend the platform's livestreaming functions to take down Periscope. YouTube's had live capabilities for years, but has focused on the Olympics and that Felix Baumgardner jump rather than the sort of point-and-stream uses that are booming now. That's going to change.
With more options and better targeting, YouTube could begin to follow you all day. "We know viewing patterns are different," Glotzbach says. "We're dealing with how we enhance our algorithms for recommendations to say, hey, this is David, we know what kind of content he likes, and we’ll recommend him this kind of stuff on his phone during the day, but when he opens up his phone—same device—in the evening, our data suggests that he’s oftentimes at home casting it or streaming it on a set-top-box-type player, so we’ll actually tweak the recommendations a little." YouTube can use your location, your browsing and viewing history, the time of day, and more to figure out what you might want to watch. Internally, they're so happy with the recommendations engine that it's now turned on automatically---you may have noticed that when you get to the end of a YouTube video, another one plays. "We’re pretty sure we’ve always got something for you to watch next," Harding says.
The team at YouTube is also helping its creators figure out what this means for them, how it might change what they make and how they promote it. No one seems to know for sure, at least not yet. That's why they're also working on providing more insight into their analytics (Glotzbach says he'd love to just give every YouTuber a quant analyst to tell them what the numbers mean, but figures that's not terribly efficient financially) and to help them start to experiment with things like high frame rate playback and 360-degree video.
YouTube’s ambitions span from the big to the small, the infrastructure to the commenting tools, but one thing is crystal clear inside the company: YouTube doesn't make stuff. It’s a platform, and the only way to win is to be the best platform. That means helping people get more views, and showing viewers new things. It means making more creators a whole lot more money. Mostly it means that whether you're using an ethernet connection on your brand-new iMac, streaming in 5K, or downloading a video via 2G in India, you better be able to watch what you're trying to watch. When that video's done, there'd better be another one waiting. And it better not buffer.
