The Exchanges

Q And while you were an undergraduate, um, and you sort of got in with Justin's kind of cohort, um, there was a guy named Matt Fong, and from, from what I understand, at, at, at a certain point in college, he, he said to you and Justin, maybe others, hey, we should start a business. Was, is that, is that what happened?

A Yeah, Matt, uh, Matt's a really good friend. He was one of Justin's roommates, uh, freshman year. Matt is my most, like, uh, Finance oriented of my friends. And I don't mean that he's interested in money. I mean, like he looks at the entire world in terms of like what's underpriced and what's overpriced and what should I be doing based on that today? Uh, and he looked around in college and he said specifically, not that we should start a company, but we have better access to intellectual capital today than we ever will in the future, uh, at a lower opportunity cost. And so we should take advantage of that, which means we should start a company.

AI assessment note: “he said specifically, not that we should start a company, but we have better access”

Q And it was his perspective, right? The camera was what would be mounted on his head, but also there'd be cameras inside the apartment.

A Mostly it was just from his perspective all the time, and sometimes he'd take the camera off and point it at himself while he was doing something, but it was his perspective most of the time, uh, walking around. Actually, one of the big things we discovered was, almost immediately, people liked him taking off the camera and talking to them more than they liked almost anything he could be doing. If he was doing something really interesting, like he's taking a trapeze lesson or something, then the point of view camera could be interesting, but 98% of the time, you actually would rather have him just be sitting at the computer talking to chat, which is, in fact, how everyone does it today, because that was actually just more compelling as content.

AI assessment note: “Mostly it was just from his perspective all the time”

Q And when you got to college, did you, I know you studied computer science at Yale. Did you have a sense of what you thought you were going to do? Like, did you figure, okay, you know, I'll get this degree and I'll go work for a tech company?

A Uh, I actually purposely didn't take any computer science because, uh, I loved programming already. Like, I enjoyed programming a lot as a hobby, and I kind of didn't want to ruin it by making it a profession or a career. And then, uh, it got to be junior year, and I'd spent freshman and sophomore year taking entirely electives, whatever I wanted, and I was like, I have nothing that resembles a major. Uh, what am I going to take, you know, another, you know, two, 11 courses in, in the next two years? And I was like, well, I know I'm good at programming, I know it's something I love. Maybe I'll try taking computer science. And I almost immediately was like, oh, this, this was silly. Like, I'm, not only am I, do I, am I good at this, but like, I really, really enjoy the, like, the work is fun. Uh, and so then I just took, you know, took, 80% computer science for the next two years and finish the major.

AI assessment note: “didn't want to ruin it by making it a profession or a career.”

Q Yeah. All right. So it's, it's 2006. You've sold, uh, Kiko for, and you guys each have about 65 grand. And did you both immediately know that you were gonna together once again start something new?

A We didn't even talk about it. It was just assumed. Like, like, I, I can't remember. I don't remember a conversation. I don't think we're like, oh, should we start another startup together? It was like, what startup will we start together? What's next? Was the only question. I think the other thing we did during Kiko that's super important to remember is like, uh, every two or three weeks we'd get, I don't know, bored, dispirited. It wasn't taking off. And we'd have some other great idea, like we had this idea we were going to build a social network for families, or we were going to build, ah, ah, YouTube, but for audio. And we would then go build and launch these products in like a, we'd like spend three weeks, and we'd do a sprint, and we'd just build this product from scratch, and we'd launch it, and then it wouldn't take off immediately. And we'd get the spirit, and we'd go back to working on Kiko Calendar. And so we probably built six startups in 18 months. I think we had this mental model that the way you had success in startups is you have this great idea, and you build it, and then you launch it, and people are like, wow, this is amazing, and they just, like, immediately it starts working. That was our mental model for, like, what success looked like. And so we just, you know, we'll just take more shots. So we just, we took, like, six shots on goal on having something th…

AI assessment note: “We didn't even talk about it. It was just assumed.”

Q Right. So just to be clear, there were other competing services that, that were doing something like this?

A Oh yeah. I mean, and, and bigger than us too. It's like Ustream was a, a live streaming competitor. They had more gaming and they were bigger than us. Uh, livestream.com. There was a European startup owned three D that was, I think already functioning at this point. Those were the primary competitors in the beginning. And so we do these little customizations to Justin TV. Um, we, we promoted the gaming content more. We made the front page promote a lot of gaming content more. And I just went out and I talked to a ton of streamers and I just wanted to understand, like, how did you get started streaming? And what do you like about it? What do you get out of this? And the biggest surprise for me was how much they wanted to make money. Um, and not because they wanted to make a living. Like, they knew they would be, I was like, you know you're gonna be making, like, 17 dollars a month, ah, given how big your audience is and how much money we can make you. And the streamers would say, like, yeah, totally, I know that. Uh, that would be amazing. I would love to make 17 dollars a month streaming video games. And I was like, okay, like, I believe you. And so one of the very first things we did is we started this partner program where you could press a button and we would roll a mid-roll on your channel and we would give you half the money.

AI assessment note: “Oh yeah. I mean, and, and bigger than us too. It's like Ustream”

Q Yeah, I mean, because you worked hard on it, I mean, you could just kill it, uh, you could presumably look for maybe somebody to buy it, but maybe nobody was gonna buy it, so did you kind of start to go down that road?

A Yeah, we, our, our thought process was, well, we don't want to do this anymore, we've built this thing, and like the, The code's fine. Like, it, it's a, it works. It is a calendar. It's a web calendar that exists. Someone must want this. Where could we sell our startup? And we decided, well, the obvious place to sell it would be on, uh, on eBay, because eBay is famous for hosting, you know, selling weird stuff at the time. Um, still today, I guess it is. And so we, we created an eBay auction to sell our startup. We just, we just made a listing that said, you know, for sale, Kiko calendar, uh, asset sale, all of this, Assets of the startup, including the domain name, the logo, all the, you know, software, uh, all the customer relationships. Uh, we will sell you the whole thing. And, uh, uh, two cows in Toronto bought it.

AI assessment note: “we created an eBay auction to sell our startup”

Q And, I mean, people knew where, where you lived. Like, you were getting, people were sending, like, hundreds of pizzas to your door and, like, calling nine one one, and the police would show up because they, they, there was a report of, like, a robbery or, right? I mean, all kinds of things were happening in your apartment.

A Yeah, we, all the harassment that streamers deal with today, we, we got in spades because we were the one, the one target. Um, I had a gun pulled on me. Um, there's a, they got a, they reported a stabbing in our apartment, and the police came in, weapons drawn, And, uh, you know, I had a, uh, uh, pistol pulled on me, uh, by the police. Uh, they, they pretty quickly realized, wait, this is a bunch of engineers, like, in their underwear programming. Like, this is not a stabbing. But, uh, that was a really scary moment. And, uh, you know, we had, we had the fire trucks called, and even if the fire trucks had showed up, we would deal with it off camera so that there wasn't a reward for the person doing it. And that, uh, it definitely, uh, over time, uh, Cut down on the amount of harassment we experienced.

AI assessment note: “all the harassment that streamers deal with today, we, we got in spades”

Q And what could the viewer do initially? I mean, could it, was it, was it passive where you were watching somebody play a game and you would see their face talking into a microphone about what they were doing? Or, or was there a chat already, um, integrated into, or you could, you could actually, you know, write your comments.

A From the very beginning on just TV gaming, all the way through when we launched Twitch and through the whole time, There was always a webcam of the person's face over the, over the video, uh, talking to the audience, because you're, you're there, you come for the video, you come for the, for the, uh, entertainment and the, the gaming content, but you actually stay for the relationship and for the chat and for the connection. That's been a core part. That was a core part of the Justin show. It was a core part of Justin TV. You can talk to the people making it, and they can react to you And you can impact what is happening on the stream in real time.

AI assessment note: “You can talk to the people making it, and they can react to you”

Q by Paul Graham and was one of the really kind of early influential incubators and is now, of course, world famous. Um, And you applied for, to be part of Y Combinator, I guess, which, and they would give you a little bit of money and some support, um, and what was the, I mean, this was the product that you were, uh, you were pitching them, this calendar product?

A Yeah, I'd been a, uh, I'd been a Paul Graham, I don't know, I was like a Paul Graham fanboy through college. I'd been reading, reading his essays. I, you know, I loved his writing. When I heard like, oh man, he's doing a incubator. I was like, this is perfect. Uh, this guy's really smart. Uh, he's had a successful startup before, and, uh, we had no idea how we would like, how to raise money or like what we would do to, we had this, we're going to start the company. And what that actually meant in terms of doing things was just Uh, other than make a product and put it on the internet, we weren't really sure, and so, uh, it seemed like this, it was this, what a, what an amazing, positive, lucky coincidence that this thing exists now. Um, so we applied, and now we applied with, with Kiko Calendar, the thing we were building. Um, the original YC was heavy on, like, show us a demo, and interviews are now 10 minutes. At the time, you've got 40 minutes per, uh, Per interview.

AI assessment note: “we applied, and now we applied with, with Kiko Calendar, the thing we were building.”

Q All right, so Matt had this idea that you guys should, should start a company while you're in college, and what was the next step? It was like, okay, let's start a company. Yes, that sounds great. Uh, what should we, what should we do?

A So Gmail had just come out, you know, summer of doesn't form. So, like, right before our senior year, I think we all looked at Gmail, and we said, this is the future of, like, software. Like, everything's gonna be like Gmail in the future. I think we could all just see that. It's, it's, it was like, uh, like having desktop software, but better. And what we did was we're like, oh, well Gmail's missing a calendar. Like there should be a, like Outlook has a calendar. There should be a calendar that goes with a email client. Um, and so we, we just started building it because that seemed like the right thing to do. And I look back at it and there's so many things wrong with the thought process we were going through there. Um, we didn't know who we were building it for. We didn't use calendars. We didn't talk to anyone who used the calendar. Um, We, we really didn't have any idea what would make a good calendar. We, we weren't, we weren't heavy calendar users.

AI assessment note: “Gmail's missing a calendar... and so we, we just started building it”

Q And how did you make people aware of it in, in, in this, in sort of the fall of 2005? I mean, most people are still on dial up at that point.

A We got, uh, we got coverage on TechCrunch. Um, we got, like, that was, which was the, which was sort of getting started around that time. Right. It was basically known for doing startup coverage. Uh, We'd try to get other press sometimes, uh, but honestly, that's a really good question that we would, we would have, it would have, the answer is mostly we didn't. Mostly we didn't get very many users. I think in the year and a half we used it, we had 90,000 people sign up for an account, which, you know, isn't tiny, but Over a year and a half on the internet, it's actually not that many, and most of them obviously didn't keep using it. And I think we had some good core instincts on how to design a calendar that worked.

AI assessment note: “We got, uh, we got coverage on TechCrunch.”

Q And what about, I mean, I hesitate to say this, because this is, like, not, you know, I don't really go there normally, but what about, like, because, you know, romantic life? I mean, what, did, did, did, was it just, how did he handle that?

A He went on some dates. I mean, he was up front with, you know, what he was doing. Um, yeah. I don't think any of them went that well, most of the time, because, like, you know, it's, it's a, like, awkward experience. Uh, but, uh, There was an incredible experience after he went on one of the dates where it just felt like it went well, and, uh, uh, he was walking home, and people start streaming out of the buildings, all of our friends lived around there, and they ever start streaming out of the building and coming up to him and congratulating him on having had a good date, and you see, you watch the date, and then you're watching that, and that was a really interesting, like, that was a cool experience. Like, there was a, you could see this, like, glimmer of something happening there, uh, that was kind of dynamic and interesting.

AI assessment note: “He went on some dates. I mean, he was up front with”

Q so that's very, it's a very different response than say, you know, some other tech companies, but, but in general, that's your position. You're like, we, we decide what people are allowed to say, what games they're allowed to play. And, ah, and how do you, how do you sort of, tell me how you kind of think about standards and, and, and, and what's allowed and what's not allowed?

A So first and foremost, I think of Twitch as being a, it's about community and about connection. And part of being part in a community is, is being willing to join that community's standards. Um, and it is not easy to set standards that work globally, that makes sense. Right now we tend to have one global standard, which I think is a good safe place to start, but we could, we can do better. And we're fortunate that unlike most social media, You can't really go viral on Twitch. It's almost, because, because it's live, it can't go, you can't get to, oh my god, look at this, millions of people see it. It's like, however many people saw it, saw it, it's over.

AI assessment note: “Right now we tend to have one global standard, which I think is a good safe place”