Emmett Shear, cofounder and CEO of Twitch, the fast-growing streaming service popular with gamers, discusses the unique dynamics of the creator economy, how it differs from gig and “W-2” work, where it may be headed, and the delicate balance platforms like his must find in creating a vibrant but safe streaming community.
Emmett Shear is cofounder and CEO of Twitch, the fast-growing streaming service popular with video gamers, which was purchased eight years ago by Amazon for nearly $1 billion. Emmett is also a part-time partner for the startup accelerator, Y Combinator.
Twitch has been in the news recently: It was the platform used by the alleged shooter of 13 people at a Buffalo grocery store to livestream his actions. The company says the stream was removed less than two minutes after the violence started, but, as Shear explained to HBR editor in chief Adi Ignatius, “there’s obviously still work to be done” to make livestreaming platforms safer.
Ignatius sat down with Shear in this episode of our video series “The New World of Work” to talk about:
- Tactics and strategies for keeping platforms like Twitch free of racist, hateful, and abusive speech.
- The present and future of the creator economy, which he says promises to be, for some, a more fulfilling alternative to not just the gig economy, but regular full-time employment.
- Monetization, and how it works differently for small creators with 5 followers, and experts with 5src,srcsrcsrc followers.
- From his time at Y Combinator, the best pieces of advice he’s developed for young startup entrepreneurs.
“The New World of Work” explores how top-tier executives see the future and how their companies are trying to set themselves up for success. Each week, Ignatius interviews a top leader on LinkedIn Live — previous interviews included Microsoft CEO Satya Nadella and former PepsiCo CEO Indra Nooyi. He also shares an inside look at these conversations —and solicits questions for future discussions — in a newsletter just for HBR subscribers. If you’re a subscriber, you can sign up here.
ADI IGNATIUS: Emmett, welcome to the show.
EMMETT SHEAR: Thank you for having me.
ADI IGNATIUS: I really want to talk to you about Twitch, how Twitch represents the evolution of the economy. I want to get into all that in a second. But before we start, it would be odd not to talk about the tragedy in Buffalo. The alleged perpetrator was initially trying to livestream on Twitch, and that has brought out criticism about whether platforms are responsible for the content that’s on them. I’d love to give you a chance to address that.
EMMETT SHEAR: It’s obviously a horrific hate crime. And that kind of violence, that kind of hate, white supremacy, racism, don’t have any place, I mean, really anywhere in our society, or especially on Twitch. And we’re working, of course, with law enforcement and taking every action we can find to respond to the tragedy. And we’re going to continue to invest heavily in ensuring the safety of everyone on Twitch. I think this is an example of one of those places where we’ve done a lot of work, but there’s obviously still work to be done.
ADI IGNATIUS: I appreciate your addressing that, and obviously this is a tough time for all of us.
Let’s get into it. In some of your comments right there, it points to the fact the world is changing and there are platforms like Twitch that are playing new roles in the economy and in how we think about business.
Creators: they’re non-employees. With your front row into this aspect of the economy, how do you see the creator economy evolving?
The New World of Work
Candid conversations on talent, tech, and the future of business. A special email series for subscribers.
EMMETT SHEAR: One of the things about the creator economy that I think is most interesting, from the point of view of streamers, is the degree to which being a creator is a career path independent of any one company. Almost every creator these days is on multiple bits of social media, right? They’re on every service.
And they may pick one that’s their primary service. One place they are most likely to go to, but it’s not like a traditional job where you would have an employer and that employer would choose the ratio of your usage. Instead, there’s a bunch of different places you can spend your time, different ways you can connect with your audience, different ways you can monetize so you can grow your audience, and you pick the balance.
We’re always in competition with every other service, trying to win the time, effectively, of the creator and convince them that we’re the best place they should be spending their time, we offer the best combination of reach, of positive interaction with their community, of monetization. And that’s a very different dynamic than a regular job.
ADI IGNATIUS: At so-called regular companies, we think a lot about culture and we think that to develop and sustain a culture you need employees who are committed to you and committed to the long-term, and you’re skilling them up and mentoring them and giving them reasons to be there. Do you think about culture differently?
EMMETT SHEAR: There’s a question I asked an employee one time who asked me a question about culture. This was many years ago. I mean, we were first getting started 12 years ago, 13 years ago, and they asked me, “What do you think our culture is?” And I asked them, “Well, what do you mean by culture?” And they gave me the best answer to that question. Something that I’ve kept as my answer to the question ever since which is, “Oh, well, your culture is the set of behaviors that you reward and the set of behaviors that you punish.”
If there’s a set of behaviors that get you positive reinforcement, you’ll get more of those behaviors in your company. And if there’s a set of behaviors that you react negatively to, you’ll get less of that. And over time, that is your culture. Culture is what we do, but people ultimately are rewarded by what causes success and what avoids failure.
I would add a nuance. My addition to that is it’s actually the set of behaviors that people perceive are punished or perceive are rewarded. If in fact you’re rewarded for doing one thing, but everyone believes you’re rewarded for doing something else, they follow what they think they’re rewarded for, not what they actually are always. So, occasionally a little gap opens up there.
So, what’s our responsibility as a service? Well, it’s pretty different from a full-time employer, right? As a full-time employer, we get all of your time. We can dictate your working hours. We can dictate all kinds of things as an employer. For a streamer, they choose how many hours they do, they choose when they do it, and they choose like what content they want to do. We have very little control compared to the normal thing.
And yet, there’s a certain set of things we can reward and a certain set of things that we can punish. Our culture is the combination of what causes your stream to grow, what causes your monetization to go up, what causes you to get discovered on Twitch. How does our discovery work? We choose which streams go through the recommendation engine, and therefore whatever we choose is the way that what tends to get promoted, that is the set of stuff that you’ll tend to get more of.
And then obviously if you act out of line, if you say something racist or hateful, if you say something that is abusive towards other streamers, we’ll punish that. Those sets of tools are how we manage creating a culture for streamers.
But actually I would say that’s just the baseline, because really a lot of the culture from streamers comes from the other streamers. A lot of the reward people get is from other people saying, “Wow, that was cool.” And so, there’s this whole second layer of culture that’s been created by our streamers themselves. That’s traditional for Twitch, actually. We might provide the baseline, but ultimately almost everything gets done by our streamers.
Our culture varies a lot by game. This fighting game culture is very different from League of Legends culture.
ADI IGNATIUS: That sounds unique to what you are as a streaming platform. Do you think more traditional companies can learn from that? I mean, do you think there’s a shift happening, not just in streaming, but where creators have the power, or gig workers have the power, that we all need to reckon with?
EMMETT SHEAR: I think bringing up gig workers is a really good place to look at, because I think gig workers are an incredible contrast to the creator economy. I don’t know if there’s a formal definition for both of them, but for me the difference is [that] the gig economy is commodified labor, working for the API. So your job is reactive to the computerized dispatch from jobs.
Which is similar to working for a creator economy company like Twitch, but where creators are differentiated, uncommodified labor: people care which creator they connect to you. You can’t just substitute one creator for another. The consumers don’t view it as the same.
Gig economy labor is commodified. You don’t really care who delivers the food, or who gives you the ride, like which exact person it is. If you got one person instead of another, it’s pretty much the same to you as a consumer, which is quite different from if someone who’s a streamer or an artist or a writer.
Are we going to see this kind of culture happen? To have a culture, you need to have a sense of “us as ourselves.” People have to know how things are done here. In the creator economy that tends to happen because all the creators talk to each other. They’re all on every social media service and therefore they’re all talking to each other and they’re all connected.
In the gig economy, I don’t see it as much in the same way, because it’s more atomized. And I don’t know if you can meaningfully speak of a culture there without a sense of collective we and us. Now you have that for W-2 employees in most companies, you have that for people who are full-time employees, because they have a sense of “we” and “us.” We all go to this place together. Although I do wonder in the future, where people are working remotely, I think many jobs might feel less that way. It might feel a little bit more like a gig economy job, a little more isolated.
ADI IGNATIUS: I probably should have done this sooner, but I would assume that some of our viewers are like, “Okay, I love this. I love Twitch. This is really cool.” And probably some people are, “Okay, this seems interesting. I’m not really sure what we’re talking about here.” Quickly, what is Twitch, how has it evolved, where is it going? To level set a little bit for people who aren’t familiar.
EMMETT SHEAR: Twitch is a live video streaming community. We enable anyone to stream live, build an audience, interact with that audience via chat, and monetize. And we’ve been known most for our gaming content. So people streaming video games, much like someone might stream sports or cooking.
But over the past few years, we’ve actually expanded quite a bit out of gaming. Our biggest category now is actually talk shows. And people just chatting, about finance or about politics or about just whatever catches their fancy. And that’s been a big category.
As we go forward, what I’m really excited about is expanding the set of people who can stream, because streaming, it’s a very socially connected form of social media. Most social media is pretty isolated. Streaming you do with your audience. You don’t do it just by yourself. And the audience does it with each other. And it’s participatory in a way that video on-demand (VOD) content isn’t. On-demand video clips are a little bit more isolated intrinsically. Live video you get to do with people. It’s interactive in that way. And that means that every stream has its own little culture. And I think where we’re headed, we want to bring that to everyone.
ADI IGNATIUS: To what extent do you police that? To what extent literally can anybody stream or to what extent do you set rule, guardrails? “You can do this. You can’t do that.”
EMMETT SHEAR: We have a pretty extensive set of community guidelines. We update them regularly. We have a pretty robust reporting and monitoring process around that, combination of proactive and reactive tools.
And, generally, I think one of the interesting things is that content can’t go viral very easily when it’s live. Because if there was some crazy viral moment, you were either there or you weren’t. It’s gone now. And so what tends to happen is [that] on live, channels build up slowly over time. So you’ll have someone who’s a bigger streamer, who has lots of people who watch them every day, but it’s hard for something crazy to happen that goes really viral. And a lot of the worst content, a lot of the things you have to really watch out for, are viral content. And so that makes our job in some ways easier, where if you’re a big streamer with a lot of reach, who’s going to impact a lot of people, you’re going to be careful not to break the rules because you have a lot to lose.
Whereas if a piece of content can go viral, people will push the limits more because if they get punished, “Well, whatever. I’m not a big user of the service anyways. Doesn’t really matter. I can just create a new account.” And I think that skin in the game for the people who have the reach is really important. Having a lot of reach, it took you a lot of time and a lot of effort to grow that reach. You do not want to lose it. And so I think that’s a big part of our enforcement.
ADI IGNATIUS: Our show, by the way, is streaming on LinkedIn, we’re streaming on YouTube, and for the first time on Twitch. So hopefully developing a new audience with Twitch on that. You mentioned that there are talk shows, things like this, but a lot of what people think about when they think about Twitch is gaming, and other things that you could view as sort of solitary entertainment, and yet people are watching. And that’s a whole different aspect to entertainment. To what extent did that grow and expand during the pandemic when we were on lockdown part of the time and maybe this entertainment was more appealing?
EMMETT SHEAR: We find that Twitch grows a lot whenever people have spare time at home. So there’s a big jolt of growth every Christmas break, right after Christmas and New Year, there’s this surge of growth then. We find that you see the same thing around other holidays on occasion, although not quite as big as Christmas. And the pandemic was sort of a giant holiday from that point of view. You had a bunch of people who were trapped at home needing time and needing connection. We grew a lot during the pandemic and we retained a good deal of that growth since. Not all of it. Some of those people have gone back to doing whatever they were doing before, hopefully going outside. But many people have kept using Twitch.
And I think that actually points one of our biggest challenges. Twitch is a lot of fun, but it’s not that easy to learn. It takes some time and commitment to figure out how it works. We tend to grow when people have that time available.
ADI IGNATIUS: Web 3.src, the metaverse: where do you see Twitch fitting into those scenarios?
EMMETT SHEAR: Let’s take VR separately from crypto because they get entangled in this metaverse thing, but they’re both pretty separate phenomena, and you could imagine either one of them being successful without the other.
So crypto, we have yet to find the thing where Twitch engaging in crypto creates a benefit for our streamers. You can imagine a bunch of things we can go build, and every time someone pitches me on one, I ask them, “Why would the streamer want us to do it instead of going to a third party? Like, how do we add value above just using one of the existing tools?” And nobody ever has an answer for me. I’m interested, but our job is to build stuff that will help our streamers. So if there isn’t any value for the streamer in us doing it, any extra value, why would we do it again? What’s the point? We’re still looking there, and I’m not sure that we’ll ever find a crypto thing, maybe.
When it comes to VR, VR is a social media platform. At some level, VR is going to be, eventually, a place people spend a huge amount of time, and as more people spend more time there, I’m sure there will be opportunities to bring live interactive streaming communities into VR and out of VR. There aren’t enough VR headsets yet. People don’t spend enough time in VR. VR also has a max of about maybe an hour or so before you run out of batteries on most headsets. Twitch is a long-form viewing service. We’re not a great match for VR today. That said, hardware is just getting better every year, and I’m very confident, as that grows, we’ll wind up doing stuff, and it’ll be fun. I think [it’ll be] really cool to just be in this video with the streamer. I think that’s going to be a very neat experience.
ADI IGNATIUS: I want to ask about monetization. Partly for people asking, “How do they monetize it?” But then, more interestingly, what have you learned about monetization? How have you evolved your business model as you’ve seen what people actually do and what they like about the platform?
EMMETT SHEAR: Monetization on Twitch is interesting because it needs to work for people who have 5 viewers and for people who have 5src,srcsrcsrc. As you can imagine, the right set of tools for someone with 5 viewers is pretty different from someone with 5src,srcsrcsrc. As you get bigger at the very large end, advertising, sponsorships, specific brand sponsorships, things like that become incredible tools because you are effectively a little mini-media company in yourself with a 5src,srcsrcsrc-person audience.
But with a 5-person audience, advertising kind of works–it’s okay, it’s not great. But direct support from your audience works a lot. Actually, we find that for those smaller streamers, “cheering”, which is a gifting from the viewer to the streamer, works best at that sort of smaller end. People just love it and want to support it.
Then in the middle, we have memberships. We have subscriptions, and people can subscribe for $5 a month. The subscription isn’t necessary to get the content. The content’s free. The subscription is being a card carrying member of the community, right? You get special badge, you get emotes in chat, little emojis. That just empirically works best for the medium-sized channels. Although all three work for everyone.
But our job is to fill in that whole range with tools. Because streaming requires you to spend many hours at it to be good at it, right? It’s a long-form thing, and you’re bringing a community together. People can easily spend 2src, 4src hours, and if someone wants to spend 2src or 4src hours a week on something, they better be making money or else how are they going to support themselves doing it? So monetization is a really crucial part of how Twitch works. It’s one of the key things we do for streamers.
ADI IGNATIUS: But if you have a successful streamer who has tens of thousands, hundreds of thousands of followers, are you saying that traditional economics—locking in those people to a contract or a commitment—are you not doing that? Is that your point that you will always let streamers have a choice?
EMMETT SHEAR: We do obviously sign some larger streamers to longer contracts, and that’s at the high end, that happens for sure. Negotiated contracts, and they have agents, and it’s a traditional media business, but that’s not generally how it works for most streamers, and that’s not the direction we’re really headed in. We really view it as basically something where the streamer’s job is to run a career as a creator, and Twitch is one of many tools they use to do that.
Ultimately, some people might sign up with us for longer terms and commitments and things like that. We don’t see that as the key essential point for how we interact with them, because ultimately they’re not going to stay if the terms aren’t great, right? Live video evaporates too fast, right? If you’re in music or movies, you could build a library of archived content and own that content and use that to have a lot of leverage with the current content creators. For us, the content from yesterday doesn’t matter anymore. We still need your active participation. So I think we’re much less focused on that kind of situation.
ADI IGNATIUS: In a lot of the questions we received from viewers, people want to know more about you personally. For example, what are your personal passions, particularly in the Twitch-type world, in terms of gaming or sports or entertainment. What are your passions?
EMMETT SHEAR: Well, I love reading. I would say reading is my favorite activity, but video games have been a close second. As I’ve gotten older, I think I’ve shifted even more towards reading, but I still love both. Currently I’m really into Hades. It’s a great game made by a little indie studio, and really, truly a work of art. I can’t recommend it enough. But I have a pretty broad taste in video games. I’ve played a lot of different stuff over the years, and I still like trying new things.
ADI IGNATIUS: I also want to ask you about your role at Y Combinator. It’s an interesting side gig. What is something that every startup should know that they perhaps don’t realize, that you could share from that experience of your startup?
EMMETT SHEAR: I’ve been more or less involved with YC over time. The part-time partner role shifted quite a bit. It used to be more hours and now it’s fewer. But I still wind up spending a lot of time with startups.
I’m going to plug my Twitter thread. I wrote a Twitter thread on this where I literally wrote down all of the pieces of startup advice that I think are important and that have hit me over the years. I thought it was pretty good and it was because it covers a lot of ground. Just to pick one of them, I think, one of my favorite pieces of advice is for startups when they ask me, “Oh, how do you hire a new head of sales?” or, “How do you hire someone who does marketing or engineering?” And it’s always some job they haven’t ever done because they’re a salesperson, they’re trying to hire a head of engineering or they’re an engineer trying to hire a head of sales. And the answer is, “Just go do the job yourself. Go try for six months or a year. And when you’ve tried and struggled and failed and struggled some more and tried again, you will understand the challenges of that job. And when you go to hire someone, you will know what you are looking for.”
It’s extremely difficult to hire for something that you have no experience with to the point where it’s almost not worth doing. You shouldn’t even try to hire for something like that unless you’ve tried it. And it’s one of the reasons I love being a startup CEO is I’ve gotten to do, over time, almost every role in the company. And for someone who likes novelty and learning, that’s an incredible gift. And I’ve gotten to slash I’ve had to, but I view it as mostly a positive.
ADI IGNATIUS: You’ve had at least a couple of successful startups, and you’re in that world. Not everyone gets it at the same pace and some people are skeptical. They don’t see it, but maybe they get there later. There’s this idea sometimes that people have to get it and go, go, go move fast. But that would leave a lot of people out of the picture who will get there eventually, who will see what you’re trying to do, who will understand the future possibilities of whatever it is, a new technology, a new approach. What are your thoughts on how to bring the most people along when we’re all processing change in a different way?
EMMETT SHEAR: In the beginning, if they don’t get it, move on. In the beginning, you want to build something that 1src people think is amazing, even if 99.9% of people don’t understand it and don’t care about it. Those people are irrelevant because someone who doesn’t buy your product because they completely don’t understand and someone who doesn’t buy your product because, “Yeah, it makes sense but I’m just not that into it,” are the same from your point of view: no sale. They aren’t a customer. And what you really care about is building something that truly makes things great for other people, that’s really amazing.
The weird paradox of startups is that if you make something that 1src people think is truly amazing, you have a better shot at something that everyone thinks is amazing, than someone who builds something that 2src million people think is kind of okay. It turns out it’s easier to grow the number of people who think something is great than it is to grow how great something is—a lot easier. And so, ultimately, the answer is the people will catch up at their own pace, and that’s fine. Don’t worry too much about that. Focus on the thing that you can control. Focus on making the thing you’re building as amazing as possible.
ADI IGNATIUS: That’s a great answer. Emmett Shear, I want to thank you very much for being my guest on The New World of Work.
EMMETT SHEAR: Thank you for having me.