The difference between checking Facebook constantly and watching Twitch regularly is like the difference between abusing drugs and sleeping, says Twitch CEO Emmett Shear.

Shear co-founded and runs the Amazon-owned live video service, which is best known for its interactive video game broadcasts. The company encourages “streamers” to treat gaming for an audience like a full-time job, offering long-form livestreams that viewers can drop into and out of, rather than short bursts of content like the type found on most social platforms.

“The way that a lot of mobile apps work is this dopamine rush,” Shear said on the latest episode of Recode Media. “Push notify, get a nugget of new information, open News Feed. Twitch is just not that. And so, I think we suffer less from a lot of the addiction, ‘Gotta press the button 19 times a day,’ kind of interactions.”

“Why does cocaine have this terrible reputation for addictiveness?” he added. “Because it’s this drug that you’re, like, doing it over and over and over again. You keep having to go back to it … I fiend for sleep every night. But it’s not a constant in-my-head-all-the-time thing. It’s like, ‘Oh. I’m tired. I want this thing.’”

However, Shear told Recode’s Kurt Wagner that he does think all tech companies need to be “self-reflective” about the addictiveness of their products. And he’s conscious that Twitch has two groups of people it needs to be worried about — not just viewers, but also streamers, most of whom can only make money if they treat Twitch like a real job, playing for four to eight hours per day. On occasion, the company will reach out to those streamers if it’s worried about them getting burned out.

“If you’re streaming Twitch an hour a day and expecting to make a living at it, I hope you’re really, really good. Because most jobs, you can’t make a living on it in five hours a week of work,” Shear said. “… And so we’re okay with people streaming a lot. What’s not good is streaming, like, 16 hours a day.”

You can listen to Recode Media wherever you get your podcasts — including Apple Podcasts, Spotify, Google Podcasts, Pocket Casts and Overcast.

Below, we’ve shared a full transcript of Kurt’s conversation with Emmett.

Kurt Wagner: This is Recode Media from the Vox Media Podcast Network. I’m Kurt Wagner in for Peter Kafka, and I’m here at the Vox Media studio in San Francisco with Emmett Shear, the CEO of Twitch. By the time you hear this episode, Twitch will have finished its annual conference, TwitchCon, but we’re taping it a week beforehand. Emmett, welcome to Recode Media.

Emmett Shear: Thank you for having me.

Thank you for being here. I wish that our listeners could see your socks. We were just talking about, you are wearing Twitch socks today.

Yeah. By the time you hear this, TwitchCon will be over, but you should come to the next TwitchCon, where there will probably be socks available.

Well, yeah. Okay. You’re right. All of our listeners have missed out on their sock opportunity, but good plug for for next year’s TwitchCon.

Thank you for being here. There’s a lot of stuff that I want to get to, but I think we should probably set the stage by figuring out what Twitch is, and I say that as someone who knows what Twitch is, and I think a lot of people do, but at the same time, I think there are a lot of people who still maybe have heard of you guys but they’re not fully aware. So let’s pretend I don’t play video games. What is your pitch to me? What is Twitch?

Twitch is a service for multiplayer entertainment. Twitch lets you come together as a community and watch other people pursuing their passions and share that with everyone else who’s watching, and interact live with those people as they’re streaming what they love.

The biggest category, obviously, on Twitch is video games. I think it’s what we’re best known for, and there’s a great a variety of video games you can stream. So video games is a pretty big subject, and it is the, I think, the largest single category of entertainment now. But we have lots of other content on Twitch as well, and the thing that really unifies Twitch across the board is the multiplayer entertainment aspect.

What do you mean by multiplayer entertainment?

Multiplayer entertainment is entertainment where you’re not just doing it by yourself. The difference between a multiplayer game and a single-player game is, do you do it interactively with other people, or do you do it on your own? And Twitch is live video streaming multiplayer that you’re doing with the community at the same time, and that’s what makes Twitch multiplayer rather than single-player entertainment.

Got it. So you’re playing a game, you’re broadcasting live on the internet. You consider me the viewer as a second player, I guess, in that sense.

Exactly. You’re part of the show too.

Okay. And how am I part of the show? Because usually, when I watch TV or a video online, I’m very much not a part of the show. I’m a consumer.

That’s right. The traditional interaction between a video producer and a consumer is the video producer makes something, and they put it out there, and then some people maybe consume it or not. Even live content’s created that way.

On Twitch, it’s extremely common for creators to be reacting to chat live and taking questions, and just cracking jokes and asking chat, “What should I do next?” And that interactivity really makes the stream happen. I actually think that on Twitch, chat is half the fun, and we often find that people show up because they want to watch a video stream, but they stay because they want to be part of this chat community.

Because they’re talking with the actual, not just other people watching, but that’s possible too, like if …

Yeah, it’s a community of people who are chatting with each other and also with the streamer.

Right.

And the streamer is part of the community as well.

And so I want to get into, this can be very big business for some of these streamers, which I want to talk about in a little bit. But before we get there, I kinda want to take a step back. Twitch is, what, four years since being acquired by Amazon? I believe roughly four years.

That’s right. Almost exactly four, actually.

Yeah, just happened. But you have been doing livestreams and stuff for a lot longer than that. I think you were one of the Justin.tv guys, which for those who don’t know, do you want to explain what Justin.tv is?

I’m happy to explain Justin.tv.

This is like, we had Justin Kan, who is the Justin of Justin.tv. He was on a Recode Decode podcast a couple years ago, but it’s kind of trippy. I’ll let you explain what this whole thing was, and when was this, was this 10 years ago at this point?

It’s almost 12 now.

Twelve, ok.

Graduated college in 2005, and Justin and I started a failed startup called Kiko Calendar which we sold on eBay. You can read the full story if you want, just Google “Kiko eBay.”

Well, give us … What is it?

We did exactly what it sounds like. We sold the startup on eBay. We ran an auction. It was great.

Oh, you literally sold it to the highest bidder.

We literally sold the entire startup to the highest bidder on eBay.

And it was a calendar startup?

It was a calendar, that’s right.

A calendar app?

Imagine Google Calendar, but before Google Calendar existed.

Okay. And you sold that on eBay. How much did it go far?

We sold it on eBay. I’m going to get the number wrong. I think $258,000.

Oh my gosh. Wow. Okay.

But I’m worried that that’s not exactly right, but somewhere around there, or it’s close.

Well, okay. Do you feel like you got the value for it?

Oh, absolutely. I feel like we had no business starting a calendar startup, had no idea how calendars worked. Had no idea, really, of how anything worked, and we were just happy to get any money out at all.

Yeah. Okay. And that’s a lot of money. Okay. So you started a calendar thing with Justin.

Right. And so then we learned from that that ridiculous stunts can get you attention.

Welcome to America.

And so we thought the next best thing to do would be to go start another startup together, and we had some bad ideas, and we had this other really bad idea which was we should make a reality television show about Justin’s life. Basically, if you ever met Justin, he’s a character, and we thought it would be great to have a 24/7 live reality show about Justin. And so they’re like, “Great.”

We raised $50,000 from Paul Graham to go do it. And we brought on Michael and Kyle, our co-founders, and we launched a 24/7 live reality show about Justin’s life, which was about as interesting as you’d imagine. It turns out people sleep eight, nine hours a day. They’re just not interesting for 24 hours a day, and we hadn’t talked to anyone from the reality television show industry, but it turns out reality TV is all about editing. The key to reality TV is taking a week of content and editing it down to the 20 minutes that were actually interesting.

Sure.

So if you take a week of content and edit it down to the week because you aren’t cutting anything, it’s really boring.

At what point did you realize, “Wow, this is not good content”?

Almost immediately. I think it took us a couple of weeks to realize that waiting for the exciting moments just wasn’t enough to get people to stick around. Because it was live, it had this great aspect to it where every now and then something really exciting would happen on stream., and there was a desire to sort of leave it running and tune in when that exciting thing happened.

It actually turns out in many ways that we were … we’d figured out what Snapchat figured out later, but then Snapchat just did it better where, with Snapchat, you would just record the 30 seconds of the thing, when the interesting thing was happening, and then you wouldn’t have the rest of the 24 hours of boring video. And then you’d share that with all your friends for a short period. Stories was brilliant. It was the fun part of Justin.tv, just condensed into a way that was actually consumable by real people.

Got it. And yet this still lasted for eight or nine months, so you realized right away it’s terrible, and yet you still did it.

We ran the show for about three months, maybe four months.

Oh, only three. For some reason I thought it was longer.

We technically left the show up for I think maybe six or seven, but we had no audio for the last half, and Justin was just going crazy. It turns out, also, being broadcast live on the internet 24/7 is really hard on you, psychologically.

Yeah, I imagine that’s not good for mental health.

And so at any rate, we realized that what we really built that was valuable was we built the technology that could let anyone do what Justin was doing, and we had all this inbound interest from people saying, “Hey, I want to stream like Justin does.” And so we opened up Justin.tv to be a service for anyone to stream live video.

And is that what Twitch became, or is that where Twitch came from?

And that’s what leads into it to turning into Twitch. It took us several years to get there, there’s a lot of ups and downs, we had to figure out how to scale the service. We had to figure out how to actually hire employees and there’s a whole learning to startup process there.

But we realized a number of years in that — this was when I was playing the Starcraft 2 beta — the only content I really loved on Justin.tv was the gaming content, and I think I had this realization that I wasn’t the only one who had that interest. And so the idea of focusing down on gaming started.

Actually, Twitch’s original incarnation was Justin.tv Gaming. We made a gaming-dedicated section of Justin.tv and started growing that, and it just got so much momentum, we realized, “Oh, this deserves its own brand and its own landscape,” because it was a different experience. But that was the genesis of it, was coming out of that Justin.tv Gaming start.

And what year was that about, that you feel like Twitch finally, or Justin.tv Gaming became Twitch or whatever, when you guys feel like you’ve kind of figured it out?

It was October 2010 when we made the decision to pivot and focus on gaming, and it was E3 2011 when we launched the Twitch brand.

I find it kind of interesting that some of the more recent efforts around livestreaming from Facebook or Twitter or even YouTube have, those are the last couple of years, so you were kind of doing this seven or eight years ago. Why do you think that Facebook, for example, launched Facebook Live when it did? What took so long?

I have, unfortunately, no idea why our competitors launch things when they do. They don’t write me letters explaining it. In fact, I think I’m on the specific list of people that don’t get informed.

Probably.

But I think live video can be a scary category for big companies sometimes, because it’s just very complicated. There’s a lot of moving parts, and you have to have a vision for how you’re going to build a community around it.

I think one of the things we’ve learned about live video online is that what makes it special is that community, at least that’s what’s worked for us. And I think building that community is really hard. So I’m not surprised that it’s taken a long time for people to take a whack at that.

Yeah. Okay. So 2011, you guys launch, we said it’s been four years, that would have been fall of 2014 you sell it to Amazon for a billion dollars, just under a billion dollars. How does that process come about?

So we don’t really talk about the method of getting to the deal, unfortunately, because there’s all sorts of pieces of paper you sign that say, “Don’t talk about this.” What I can talk about is why I thought Amazon would be a good home.

Sure. Well, I imagine there were other people who were interested. Was this a formal process? Can you give us that?

I think that it’s a matter of public record that we were represented by Qatalyst, so clearly there’s some kind of process there.

Sure.

I was really excited about Amazon for two reasons. One is that during the process, I got to meet Jeff Bezos, and I was super impressed by his questions and by his level of insight and his ability to ramp up quickly on the information. And then the other part of it was, I just had to look at Amazon’s track record.

Amazon is almost unique in large tech companies in retaining founder CEOs after the acquisition. Their hit rate is not 100 percent, but Zappos, IMDb, Audible. It’s kinda crazy, actually. Most tech companies acquire the company and they just assume that the founder is going to be gone a year or two later, and I didn’t really want to do that. I wasn’t interested in that outcome. And so I’m really glad we wound up going with Amazon because it’s worked out for us, too.

Was there a specific vision that he had or conveyed to you, like what was the general pitch? Because you do have options at that point. So what is his pitch to you?

The Amazon pitch, I think, for acquisitions, at least at scale, is a really simple one, which is, we’re going to be a good home for your business, I’m going to let you run it, and we think that you’re the best person to run your business, we think you know more about your business than we do. And we’re buying it because we want you to keep doing that. So if you’re not interested in that outcome, please don’t take our money. That would not be a good outcome. And that was appealing to me. That’s what I wanted, was independence and the ability to keep running Twitch.

And it’s been four years, and you’re in San Francisco.

I am.

Amazon … is the whole team?

The whole team is here too, yeah.

So your whole team’s here.

Well, we have people in other offices as well.

Headquarters.

But a great majority of Twitch employees are in San Francisco.

So do you feel like that autonomy, how has that manifested over the last four years? It seems like if you’re still here, that must mean that you feel like they’ve lived up to their end of the bargain.

Absolutely. We’re part of Amazon, and so there’s always issues being part of a larger global organization. A really obvious example would be tax. You have to think about the fact that you’re part of a big company where your actions could impact where you get taxed and where you don’t get taxed. That’s something we didn’t have to think about before we were part of Amazon.

But the trade-off is we get to launch Twitch Prime, and Twitch Prime is an amazing program that we could never have built and launched on our own, and it’s awesome for our streamers.

That’s the subscription program you guys have.

Yeah, it’s a subscription program, it’s a partnership with Amazon. If you link your Twitch account with your Amazon account and you have Amazon Prime, you get all these great on-Twitch benefits, you get free game content, you get a free channel subscription. And those benefits and the scale of that program is something I’m not sure we could have ever figured out how to put together on our own.

Yeah. Okay, I need to send us to an ad break, but one final question before I do: Twitter. I heard you guys almost sold to Twitter at one point.

I can’t comment on rumors.

Okay. Fair enough. We are going to go to an ad break real fast. We’ll be right back.

[ad]

Okay, we’re back with Emmett Shear, the CEO of Twitch. Can’t comment on rumors, but hopefully you can comment on the content that is on Twitch today. So you’ve talked that most of what you’re known for is gaming. So someone’s playing video games, other people are watching that person play video games. What kind of other stuff is big on Twitch today?

Well, the flashy stuff that’s on the site right now that probably is getting the most attention, that is really exciting, “Thursday Night Football” from the NFL, on every Thursday, a Pokémon marathon going through all the Pokemon episodes.

Today I saw that, just a few hours ago.

With the audience, yeah. Pokémon’s on right now. We recently did a “Doctor Who” marathon. I’m super proud of our work breaking Bob Ross back into the mainstream.

The painter.

Bob Ross was sort of unknown, he was a PBS painter, and we ran a big Bob Ross marathon, sort of a nostalgia event, and Bob Ross has just blown up. You see him all over the internet now, and I think that’s great because Bob Ross is such a big part of my childhood, and I’m glad that kids again are getting to interact with Bob Ross.

And so how do you decide, because that is not necessarily the kind of TV that I would carve time in my schedule for usually. Maybe a marathon. Maybe if you’re a diehard “Doctor Who” fan, you might say, “Oh man, there’s a marathon going on, I want to get there.” But it’s not the “Game of Thrones”-type of content that I’m going to say, “Hey, every Sunday night at eight pm, I’m sitting in front of my TV.” Do you want that kind of stuff or are you okay picking the content that’s maybe more, I call that background or fringe-type stuff, but how do you think of it?

I don’t think of it as fringe. I think of it as fandom. What unifies Pokémon or “Doctor Who” is it has those fans who just love it. And what makes the Twitch TV experience so different from normal TV is you’re together with 10,000, 20,000, 50,000 other people participating and watching this together as a community.

That’s interesting and that’s exciting to you if you want to be part of a community around this content, and so it’s important to have, as a seed of that community, people who really love the show, and then other people come in and can sorta join them, and it can expand the community, and grow the community.

I spent a lot of time watching our “Yu-Gi-Oh” marathon. I’d actually never watched “Yu-Gi-Oh.” “Yu-Gi-Oh” was a little bit …

What is “Yu-Gi-Oh?”

It’s a card game. It’s sorta like Magic: The Gathering but there’s a TV show about it. I found myself really kinda getting into the “Yu-Gi-Oh” marathon even though I’d never really liked “Yu-Gi-Oh” before, because the community was so welcoming and it was such a unifying experience.

Just because people were talking about the show while you’re watching the show?

Yeah, you can participate. People are kinda making fun of it sometimes or they’re laughing along with it. They’re making kinda memes out of the events of the show and it’s fun. It’s another layer of experience.

How do you decide what shows have that fandom necessary to appear on Twitch?

We ask our audience. We’re all about serving and talking to Twitch people and finding out, what do they like? What would they like to watch a marathon of? We use that data to figure out what’s gonna appeal.

What is something that maybe you would love to have on Twitch that you don’t today? Is there a certain show that you’re a big fan of that’s not on there?

I mean, I think that the main thing that we look for is something that will be an event. We’re looking for something where, when it goes on, everybody is going to want to tune in and come together around that. And so, I think the best things there, the ones I’ve enjoyed the most of and things like are our Mr. Rogers marathon where … I love Mr. Rogers.

Isn’t there a movie coming out?

There is a movie coming out. That would actually be a great occasion to redo a Mr. Rogers marathon around that movie and have a buildup to everyone being able to go see that. I actually think the idea would be you turn off the Twitch stream at midnight when the movie starts, and everyone leaves Twitch, and they go to the movies together. I think Twitch is about that community thing.

Everything doesn’t have to happen on Twitch. We’re trying to create that community, trying to bring people together, and whether or not it’s on Twitch or off Twitch, I think that’s a win.

What about getting the movie on Twitch? Would you like to do that someday?

Movies can be fun on Twitch. We’ve done a couple movie premiere things on Twitch. The way live video works, everyone’s not necessarily available to tune in at the same time, and so we generally find longer more like marathon-style content works better, but we’re always open to experimenting. I think it would be cool to have a marathon that’s capped off by a movie release. We’re really focused on that sorta bigger event, the longer event.

How does that work with the NFL you mentioned? Amazon has this digital-streaming deal with the NFL for “Thursday Night Football,” you mentioned the games are on Twitch or also on Prime. Sports is one of the things where you really have to watch as it’s happening. How does that manifest given … to your point, right? Not everybody is necessarily free to tune in at certain times. Does football work on Twitch?

Yeah, football does work because people have already built their lives around when the football games happen.

Got it. So the NFL is strong enough that it works?

Yeah. If you have appointment-viewing content, people are used to consuming it in an appointment-viewing way. Sports, news, that kind of daily content, I think that can work. Honestly, what really makes the NFL work on Twitch is co-streaming. We have a streamer co-streaming the NFL.

Sorry. What does co-streaming mean?

Co-streaming is where you add your voiceover to what’s happening in the action. That attracts a whole new audience of people who are fans of that streamer and brings them into the football content. That’s been super successful. It’s been really cool to watch Twitch communities come together around watching the NFL the way they might’ve come together in the past where I’m watching a video game.

We’ve seen that’s been exciting for the community. I think it’s a really cool opportunity for any sports league to be able to reach a new audience of millennials and gen-Z people who might not already be huge fans. I think it’s a natural fit.

How does the content that’s non-gaming … so there’s people who are streaming from their living room or bedroom, or wherever it may be, then you have stuff like the NFL or TV shows. Who cuts those deals? Do those come to you because they’re available on Amazon Prime and you kinda get a benefit of that, or are you guys kinda going out and cutting your own partnership deals with people? How does this actually show up on Twitch?

For content acquisition it’s a mix of a whole bunch of different ways of going after it. Sometimes things are inbound, sometimes we’re going and getting it, sometimes we’re working in collaboration with Amazon to go get it. I think we’re open to a variety of ways of going after it. I’m not sure there’s any one primary way.

Definitely, most of the stuff we do is exclusive to Twitch. The NFL is one of the few things we’ve done that’s been a joint Prime-Twitch streaming program but it’s worked really well, so we’d love to do more in the future.

I was just gonna ask. There’s some good original stuff on Prime. It seems to me like there’s not a huge risk of putting it on another Amazon-owned property. Is that something you want to do? I mean, there’s a lot of Amazon originals that seem to be doing well.

Yeah. I think just throwing video onto Twitch at random tended not to work very well, but we’re always looking for ways to work with Amazon and bring Amazon assets to Twitch, and we’re always looking for ways to introduce Twitch customers to cool video, so that’s an easy match for us to go make.

How would you decide then? I’m sure to your point it’s not as simple as, “Hey, here’s good video that works on Amazon, let’s throw it on Twitch,” but you gotta also experiment or figure out some way, so how do you do that?

One of the Twitch company values is experiment to decide. I put that value in place because I really believe that the best way to figure out if something works is often not to spend a bunch of time theorizing about whether it might work or not, but try it and see. That’s how we would go and investigate this as well and I think that’s how we are investigating this. “Thursday Night Football” is a great example of experiment to decide.

We’re giving it a shot. We’re measuring it. We’re seeing how it performs. We’ll make decisions again in the future based on the data we get out.

Okay. Let’s talk about gaming, though, because that is still certainly what I think of you guys for. Can you quantify at all how big gaming is compared to maybe other types of content that you guys have? It’s the dominant thing, right?

Yeah. Gaming is the super majority of our content. It is the primary thing people come to Twitch to watch today.

And so, what do you feel … I was looking at some of the big games, some of the obvious ones if people are gamers they would say, duh, is Fortnite, League of Legends, I think Call of Duty was up there. What’s the big draw for you guys today?

We do well with all basically multiplayer competitive games and we also do well with games that I would call sorta sandbox, or creative games. Twitch tends not to get a ton of broadcasting for more single-player experiences. I think sorta unsurprising were multiplayer entertainment, and single-player kinda takes you away from that. It’s the multiplayer games that do really well on Twitch in general.

Which is Fortnite. Explain what Fortnite is?

Fortnite is huge for us. Fortnite is a hard game to explain. You’re parachuted onto an island where you pick up weapons and can chop wood to build structures to defend yourself while you’re trying to defend yourself against 100 other people who are also on the island and be the last person not to die.

Right. You’re all shooting at each other.

If you haven’t played a Battle Royale game — I hope everyone listening to this has played a Battle Royale game — if you haven’t played a Battle Royale game, you should go play one.

I’m gonna assume that people haven’t played Battle Royale.

It’s pretty popular.

I’m not saying it’s not popular.

Maybe I’m just surrounded by a bunch of computer nerds who all play that.

I’d say you are immersed in the gaming community more than probably most people who will listen to this.

But yeah, there’s a whole genre now. Fortnite is extremely prominent but there’s Playerunknown’s Battlegrounds. There’s a new Call of Duty Blackout title … and the joke at E3 actually was that everything was a Battle Royale game this year. There’s a ton of I think actually really exciting, cool, innovative things people are doing in the Battle Royale genre and they’re all around this idea of, you and 99 other people show up on an island, who is the last one left alive?

Right.

Or sometimes you have teams of two, teams of four.

How are people … streamers are making a lot of money doing this, some of them are, walk us through a little bit about how that actually works? So I’m a great Fortnite player, maybe I have a good personality too, therefore people want to not only watch me play video games but watch me kinda narrate as I’m playing video games. How am I making money from that?

People definitely love watching people who are both good at the game and entertaining, and if you can hit both of those you can get quite popular. You make money in a variety of different ways when you have that audience. You can show advertising, which is a tried-and-true method of monetizing an audience that I’m sure listeners to this podcast are familiar with.

I’m sure they’ve heard of it.

I’m sure they’ve heard of it.

We’ll assume so.

We also sell channel subscriptions. A channel subscription is something that the streamer can sell that costs $5 a month that gains you access to a special set of emoticons you can use in chat, a special badge next to your name, the ability to speak when the chatroom is put into subs-only mode. It’s not access to the content; anyone can watch the content for free, anyone can watch the video for free. It’s more like joining a fan club. It’s like being a member of a club and really supporting the streamer.

One of the really cool things that you get with Twitch Prime is a free subscription every month so that’s a great way to build a base of true fans. We also have “cheering,” which is based around using Bits, which is this virtual good that you can use to create a celebration of a special moment.

I think of them as tips. I mean, you guys call them Bits.

Yeah, you could use that metaphor. Bits are like a virtual good. They’re like these little gems and you can use them to make explosions and different kinds of emoticons on stream.

Does it surprise you that it has become such a big business for certain people? How much are people making I guess? Give us a sense.

People can make hundreds of thousands of dollars, millions of dollars, on Twitch over time. I think that we’ve really seen that entertainment over the internet isn’t really any different from entertainment off the internet. The best talent and the best creators, of course, can make money at it. It doesn’t surprise me at all. In fact, it’s literally our company mission, is to enable creators to make a living entertaining and educating their fans.

Was that something that you wanted to do when you started? You said you were a gamer.

Yeah, that was actually the goal when we started Twitch. When we pivoted Justin.tv to Twitch, I realized we needed to make a decision as to which audience we were gonna spend the most effort on. Were we going to be focused first on helping creators make a living, entertaining and educating their fans, or are we going to be focused first on helping viewers find entertaining creators they can watch? Obviously, it’s two sides of the same coin, but picking one lens that’s primary I think is really important, especially for a startup.

We choose streamers because we think that’s the heart and soul of what makes a service like Twitch work is the creator.

And so, one of the most famous creators is this guy named Ninja, right?

NinjaTyler, yeah. Ninja’s awesome. He’s been on Twitch since before Twitch was Twitch. He started with that when we were Justin.tv, I think, so he’s been on for a long time.

He’s the most famous guy, right? Or at least the most successful.

He’s certainly the most famous Twitch streamer right now, yeah.

And so, explain to the listeners what happens, because he went on with someone famous like a musician, right? Didn’t he go on with Drake or someone like that?

He had Drake on his network. He didn’t go on with Drake. No, no. Drake went on with him.

Sorry. Correct. Drake was the guest.

Let’s be clear. I had this confused look on my face that you can’t see.

I know you did. I was sitting here like, oh my God, did I get the name wrong or something like that?

I was like, he didn’t go on any musician’s stream. That never happened. No.

Correct.

Drake came on Tyler’s stream because he needed to reach a new audience, I think. He needed Ninja to help him get with the young kids, I think.

And so, what was that like? That was a massive internet … that was one of those internet moments where people kinda realize, “Oh my gosh, this whole notion of watching others game is not just a gaming thing, it’s a pop culture thing.” That’s how I took it.

It’s really interesting because when you look at the stats, I mean, Ninja’s been on Twitch for a long time, and he’s been sorta growing steadily, and when Fortnite launched, we kinda hit this inflection point because Fortnite was so popular and he was the best Fortnite streamer, and so he just like kinda exploded in popularity, and that’s when Drake went on.

You’d think that Drake going on was this inflection point, but if you actually go back and zoom out and look at his popularity growth, Ninja’s popularity was zooming up anyways, like, that was happening because people loved Fortnite and they loved his stream, but the Drake moment was super important for the general public because it was for a lot of people the first context where they kinda heard about Twitch because it got talked about a lot.

Right, because how many viewers were there?

650,000 roughly.

Yeah, at one time, which is pretty impressive.

What was interesting about it is you might think that’s a bunch of Drake fans showing up for Tyler, but no, that was Ninja’s whole fan base showing up, because like, “Oh, this is awesome, he’s streaming with Drake.” And so that was his fan base turning out for him.

Was that something that you guys set up? Are you involved in that in any way or is that just like Drake and Ninja, Tyler … I keep calling him Ninja, I guess whatever, that’s a cool name.

Ninja’s a cool name. We can call him Ninja.

Yeah, so is that just Drake and Ninja getting together with their agents on the side or are you guys playing a role in this?

We didn’t actively set that up. I got a bunch of text messages from my friends being like, “Dude, do you know Drake is on with Ninja?” And I was like, “No, actually I didn’t,” and then I went and turned Twitch on and there it was.

You were one of the 650,000.

That’s right.

At the very beginning of the conversation I mentioned Facebook and my question at that time was what took Facebook so long? I guess I’m curious when you look at the landscape today and look at where you guys are … I think of Facebook as a logical competitor. Where do you see that you fit in?

It’s interesting. I certainly think that it’s important to keep track of your competitors and know what they’re offering and make sure that they didn’t come up with some awesome idea that you didn’t come up with.

At the end of the day, we run Twitch in a very streamer-focused way, not a competitor-focused way. I hope we fit into the landscape as the most streamer-focused multiplayer entertainment service in the world, and I think that’s what we strive to be.

What does that mean, exactly? Because you’re still … Whether you want to pay attention to Facebook or not, they’re there. Right?

Mm-hmm.

And I’m using them as one example. But I guess I’m curious what being streamer-focused versus competitor-focused entails?

Yeah. So, competitor-focused is where you spend your time worrying about how do we beat their feature set? How do we make our product stand out from that product?

Right.

That’s about you and about the competitor and it’s not really thinking about the streamer at all. Streamer-focused is like, okay, what do streamers care about? What do streamers want? Streamers want really high-quality video. They want to reach new fans. They want to make a living and earn an ability to support themselves doing it. They want something we call love. They want positive social interaction and reinforcement — from Twitch, from their fans, from their community. And you can also focus on, how do we deliver more of that? That’s where we spend our time.

We spend our time thinking about, how do we improve video quality? How do we deliver more fans? How do we deliver better ways to earn a living? How do we deliver better, positive social interaction? We think that as long as we do that, as long as we deliver as much of those as we can, streamers will want to choose us.

We don’t really have to worry too much about anyone else because we’re doing the things the streamers want, and we’re gonna do that as best we can. And as long as we execute well, that’s all we have to do.

Is it competitive to get a streamer on to the platform? I know, for example, when Facebook first started doing this, they came out and they said, “Hey, look at all these streamers that we got.” And I imagine that they offered them some kind of sweet financial incentive, to stream on a totally new platform. Do you guys feel like you have to, almost, compete for these types of people?

Absolutely. We’ve been competing for streamer attention from Day One, I think, in much the same way that Amazon is in an intensely competitive industry — there are thousands of people who sell stuff — Twitch is in an intensely competitive industry. There are a lot of services that offer you livestreaming, and we’re really honored every single time a streamer chooses us over a competitor, because we really do think there’s a choice you’re making, and we’re glad that we are that choice, and we realize we have to re-earn that trust every day.

Do you guys bring people on under contract, or anything like that? Or is everyone just, they’re free to stream somewhere else whenever they want?

So, we definitely contract with people. That’s how you get paid, is you become a 1099 contractor.

Got it.

So, on some level, it’s definitely in contract. Right? We’re not gonna cut checks to people if we don’t know who they are and don’t have signed paperwork. But I think what you’re really asking is how do we think about contract duration?

Mm-hmm.

The answer to that is a lot of our contracts are under NDA with a partner, and so we can’t comment on the exact terms.

Okay.

But I like to think that we aim for creating contracts that are fair to both sides, reasonable. We don’t try to force streamers to give up their future.

Yeah. And I should have asked this back when we were talking about the business element, but you mentioned three ways. There’s the subscription, there’s ads that can appear, and then the Bits. Kind of that tipping.

Mm-hmm.

Is all of that in some kind of revenue share element? So, the creator gets a portion and Twitch gets a portion?

That’s right. We have a rule, actually, that we try to make money with our streamers. And so, we’re always looking for things where, if Twitch is making money out of something, we want to make sure that there’s some way that’s benefiting the streamer, as well. And I guess vice versa, right? We want to find a way to benefit alongside our streamers, because that’s the way we stay in business.

Mm-hmm.

And so, we like to set it up so it’s win/win.

And so, how much do you take?

We don’t have any, like, globally standard public set of revenue shares.

That’s a secret number.

It’s like contract terms and things like that. It’s one of those things, I guess … You know. Negotiated, sometimes per partner.

Okay. We were talking about things that maybe, what I feel don’t necessarily work on live video. I am thinking of Meerkat, Periscope, early days of Facebook Live, when it was like, “Hey, go live from your living room.” It seemed like that didn’t really work for live video.

I actually have a lot of experience with mobile live video, because Justin.tv launched one of the very first mobile live video apps. And so I got a lot of firsthand experience with trying to do live video streaming for mobile. That’s why I’m so impressed with how Snapchat tackled that.

Yeah.

Because it turns out, the primary thing you need to make live video work is the ability to pull in an audience. This is still the thing that is the hardest thing for us to help our streamers with. Because when you just go live, it takes a long time to build up an audience. Unless you already have people who know to tune in at this time for you, it’s hard.

The way to build that audience is to start streaming at the same time every day, or to have the ability to push-notify a whole crapton of people all at once. And to have them not be upset that you just push-notified them so they turn it off.

Right, exactly. Not get annoyed.

Exactly. Not get annoyed. Twitch has taken the appointment-viewing approach. So, you know, we have push notifications for things, but that’s not the primary way we do things. We address building an audience by getting people to enjoy the service enough that they just show back up again, because that’s when they know you’re on, 5:00 p.m. every Friday or whatever.

Right.

Whereas, I think there’s sort of a different approach. You can try to do the push notify thing, but it’s really hard. The Justin.tv mobile app, what we discovered was almost no one ever tuned in while the stream was live. They almost always watched it hours after the fact.

Why don’t people … Is it just the content’s not interesting? Like, why didn’t that really work? I know obviously there is the audience element, right? Like, no one wants to perform for the camera to nobody. But at the same time, it feels like it … I don’t know. I don’t even have interest in really watching my friends do it, so I’m curious if it’s just bad content.

The key for, I think, all mobile creation of the short-form variety is that if you try to do it live … If you create a piece of short-form content live, people have to tune in right when you’re making it. If it’s a 30-second piece of content, I have to go watch that immediately. I don’t know how you use your phone, but I’m not just sitting there waiting for something.

No.

And so, that’s how Twitch approaches it, is we actually think we’re about long-form content. Like, one of the reasons why we have the No. 2 highest watch time in the U.S. is because it’s long-form content that you really sit down and you engage with over a long period of time. That’s really what works well with live. Trying to do short-form snippets — go live from this moment, go live from that moment — just doesn’t work as well as you might think it does. Because people aren’t, like, on call to tune in at your demand.

Yeah. I’m glad you mentioned the watch time, because I wanted to ask you. There’s a big conversation now in the Valley, of course, around tech addiction, making sure that … Facebook, for example, has started to say with Facebook and Instagram, “Hey. Here’s how much time you’re spending in the app.” We obviously saw that with Apple. How do you think about that? Because your business is similar to theirs. Like, the longer people pay attention, the better it is for business. So, do you guys think of the fact that, “Oh, do we want people sitting in front of the screen for hours on end?”

We do think about it. I think that for Twitch, because we’re not building an addiction loop, it’s a little different. Right? Like, the way that a lot of mobile apps work is this dopamine rush. Push notify, get a nugget of new information, open News Feed. Twitch is just not that. And so, I think we suffer less from a lot of the addiction, “Gotta press the button 19 times a day,” kind of interactions.

And Twitch people don’t open our app that many times a day. They just open it for a long time when they do. And I actually think that’s a much healthier way of interacting with your technology, and with entertainment. You know? I go to a movie, I’ll spend two hours watching a movie. That doesn’t mean that I’m addicted to the movie. It just means that it’s a good piece of entertainment I decided to go sit down and watch. And I don’t necessarily think that’s such a big deal.

That said, we definitely are paying attention to this, and I think that it’s really important for every tech business to be self-reflective about, you know, “Is this addictive? Are people actually getting value out of their usage?” And making sure people feel like … You know, you give them the tools to have control over that.

Yeah.

To not get sucked into the never-ending addiction spiral of phones.

Yeah. So, it’s interesting. It sounds like volume kind of matters, from your perspective. Because I’m sitting here thinking — at least I was 30 seconds ago — that whether or not I open my Facebook app 50 times within a two-hour period, or if I open it once, but I’m looking at Facebook for two straight hours, that, to me, feels kind of the same. You see those as maybe different.

I see them as very different. Why does cocaine have this terrible reputation for addictiveness? Because it’s this drug that you’re, like, doing it over and over and over again. You keep having to go back to it.

Sure.

You’re fiending for it over and over again. And, you know, that’s true for sleep as well. I fiend for sleep every night. But it’s not a constant in-my-head-all-the-time thing. It’s like, “Oh. I’m tired. I want this thing.”

Yeah.

And I think that’s actually healthy. That’s fine. Obviously, there’s a dynamic range there. If someone sits down and is playing something for eight hours a day or watching it for eight hours a day and it’s disrupting their ability to be productive, that’s a problem.

Sure.

You can get addicted to anything.

Yeah.

You can get addicted to food. But, it’s pretty different to sit down and engage with something for two hours, have an enjoyable experience, and be able to put it down and go on your way and live your life. Versus something that you feel the need to … it’s constantly you’re checking it, you’re checking it, you’re checking it. You can never put it down.

What about the creators themselves? Because as a consumer, yes, I can check in and say, “Hey, I want to watch this for 30 minutes,” and then I go on my day. For people who are … You know. This is, for some of them, their livelihood … I imagine there’s pressure to always … They can’t make money unless they’re live, right? So, do you do anything about that? Or is that on them to figure that out?

Yeah. Thank you for asking that question, actually, because I think that the streamer lens is equally important, and actually it’s one that we think about a lot. The first thing about that is that we do regard Twitch as a full-time job. Like, if you’re doing Twitch eight hours a day, you’re doing it right as a streamer. Like, you’re supposed to be streaming a reasonable amount of time. And you don’t have to do eight hours. You can do six hours, or four. Right? But if you’re streaming Twitch an hour a day and expecting to make a living at it, I hope you’re really, really good. Because most jobs, you can’t make a living on it in five hours a week of work.

Mm-hmm.

So, we’re not necessarily upset with streaming when it happens for eight, what would be an outrageously long length of time for entertainment. But when you have the context of, this is meant to be earning a living or at least earning substantial income, it seems less crazy. And so we’re okay with people streaming a lot. What’s not good is streaming, like, 16 hours a day. Right? I mean, if you do it once in a while, that’s fine …

Do you set limits?

We don’t right now, because we don’t have a way to do it in an automated way that we think is fair and good. But we will talk with people if we think they’re overdoing it. And I think that it’s an area we need to actively invest more. I actually don’t think we do enough here yet, but I think that we’re very much invested in our streamers’ long-term health.

Yeah.

We’ve had — I think Ninja’s a great example — streamers who have been on the platform for, you know, some of them coming up on eight, nine years now. That’s the kind of relationship we want with our streamers, and so we think about this problem a lot. We do not want people burning out and dropping out.

Yeah. I have one final question for you. We were talking about the autonomy you get with Amazon. You guys are down here. I cover Instagram. The Instagram founders just left, and it was like a big thing. Eventually they just ran out of autonomy, I think, at Facebook.

Mm-hmm.

What is keeping you at Twitch at this point?

What do you want out of a job? At least what do I want. What do I want out of a job? What I want out of a job is, I want something that is challenging, where I feel like I’m always learning and growing. I want something that I think I am making a difference in, where I can tell that my work has an impact advancing the thing that I’m working on.

And I want the thing I’m working on to matter. I want to feel like I have a purpose and that it matters. Twitch is checking all of those boxes. I care about it. I think we’re working on something that’s really meaningful. I feel like I contribute a lot to it being able to be successful, and I can help on that front. And I’m being challenged and growing by it all the time. Amazon has given me a playground where I can grow, and I feel like I can make the decisions I need to make and grow the company. It’s like, why would I leave?

And, you know. Who knows if I’ll be at Twitch forever? God, if you’d asked me 10 years ago, “Would you still be at Twitch?” I would have told you, well, “First of all, what’s Twitch?” Because it didn’t exist yet. But, like, “I have no idea.” Right? “Probably not.” I would have said, “Extremely unlikely I’m still working at this company.” And yet, here I am.

Yeah.

So, apparently I’m bad at predicting how long I’ll be places.

Is that autonomy a necessity for you, like it was for the Instagram guys?

I think that any CEO founder who’s running something probably needs that autonomy, and I think you can judge which companies grant that autonomy by how long they retain CEO founders. The fact that you’ve got Tony still at Zappos says something. That’s not an accident. That doesn’t happen at random.

Yeah. Emmett, this was awesome. Thank you so much for being here.

Thank you for having me.