Life At A Startup, Part 2

Part 1 is here.

I left justin.tv in 2011, just after the SocialCam and Twitch products had been launched. I felt strongly that Twitch was probably onto something, but it seemed like there wasn't much room in the company for me to grow any more. At the time justin.tv still had four founders, and it often felt like every project that needed strong leadership would naturally be led by one of them. I had immensely enjoyed the first couple of years of justin.tv, when there were few enough people that there were still big things even an individual contributer could be responsible for. But we had grown beyond that point, and I was bored and ready to leave.

A former coworker offered me a job as CTO at a startup he'd left justin.tv to found. That seemed like just the kind of thing I was looking for, but after nearly a year it was clear it hadn't worked out. Our personalities didn't mesh well, and I got quite sick towards the end of my time there, I think perhaps in part from the stress of continually grinning-and-bearing a situation I had grown to really dislike. In the space of three months I had pneumonia, a stomach ulcer, and a norovirus. After all that I knew it was time to change something!

My wife will do just about anything for her family, and decided we needed to get me into a situation where I could relax and take it easy for a while. We bought a nice house in Sacramento (mortgage $900/month, vs our $3500/month San Francisco rent!), and I settled into reading in the hammock and repairing and upgrading the house. We didn't have a huge nest-egg saved up, but it was enough that I could take a few months off without worrying where the money was going to come from.

I did get bored quite quickly though, and worked on a bunch of things just for fun. One of the most enjoyable was a hardware project. I took a bunch of Beaglebones, and some stripped-down webcams, and built a panoramic video camera. It's an idea that I've since seen a bunch of different times, and I'm sure someone will soon bring it to market. So far they're all missing the part I was most excited about though: I don't think anyone actually wants to watch panoramic video in the kind of player people build for these things, where you have to turn your virtual head to see the action. Instead of doing that, my software made good guesses about the camera angles and zoom levels a human camera operator would choose (e.g. using motion and face tracking), and just gave you a regular video file to watch.

I count myself as lucky in the extreme, that a hobby I had as a 9 year old boy turned out to be a career in such high demand that I basically never have to worry about getting a job. Sure enough, after a few months went by and I was starting to think I should stop spending our savings, a job was easily found. It was a teleworking gig I could do from my little shed at the back of our yard. It feels pretty amazing to "commute" to work by strolling down a garden path, often still in pyjamas. Every situation has its downsides though - being contract work, I had no say at all in the direction we were taking. I also missed being around coworkers.

So, when my friend Emmett - now CEO at Twitch - sent an email around looking for an engineering lead to join justin.tv, choose a new vertical, and try to replicate some of Twitch's success, I really felt like I was in the right place at the right time. In the space of a few weeks we had put our house on the market and were back in the Bay Area.

I spent a lot of time thinking about what might make a good vertical to attack. More than anything, I wanted to understand what had made Gaming such a good vertical.

At first I jumped to the obvious conclusion: Gaming was the content that was most like sports, without actually being sports and having that content's copyright problems. I just had to find whatever was the next most sports-like content. I excitedly told my wife I had it all figured out, but she didn't seem as impressed as I thought she would be. She thought Gaming had succeeded because it's something people already do at a computer, and so broadcasting it was less work compared to broadcasting other things.

I shamelessly cornered Emmett at his wife's birthday party and picked his brains. Why had Twitch succeeded? He surprised me with yet another opinion: It's the interactivity. People really enjoy watching someone do something impressive, and then being able to chat with them about it.

So I mulled all of this over for a few days and, me being the kind of person I am, I have to admit a lot of the time I was just wondering why neither Emmett nor my wife could see the obvious truth: Live video is only interesting for sports, or sports-like content!

I managed to break free of my need to be right, just long enough to realize I wasn't even close. Late one night in the office, I was randomly flicking through channels when I found an artist sketching. I stayed long enough to become quite engrossed, and was actually sad to have to leave the office and stop watching. As I rode the BART train, I was starting to think watching people sketch could work. I used to watch hours upon hours of Tony Hart as a kid (apparently the American equivalent is Bob Ross), so I could see there might be some amount of broad appeal to this kind of thing. And, though this was the furthest thing I could imagine from "sports-like", it did match up very well with the facets both Emmett and my wife had been talking about. It was a little annoying to realise I might be wrong, and my wife and my friend probably both had much better explanations than I did!

I got off the BART, and started cycling home. As I pedaled up the hill, my mind wandered and I thought about other kinds of video content, when I remembered the only videos (except for Netflix) that I've ever paid for: A series of tutorial videos on how to make music with Ableton. Then it hit me: What if we didn't just do sketching, but all kinds of creativity?

I spent a lot of the next day scouring justin.tv for any kind of content that might fit. And I found tons of it. The content existed, it was just scattered all over the place. People were sketching and producing music, as well as doing things like making jewelry, modeling in clay, blowing glass, and streaming themselves writing code. It didn't take long to convince myself, and then the rest of the team, that this was something worth persuing. We quickly made a new category on the site, and got to work trying to build a community around it.

It's been only a month since we started working on Creativity, and already I know we're on to something. I've seen an artist streaming some amazing 3D modeling work, only to get hired on the spot by a viewer. Another guy is building a car literally from scratch, milling all the pieces from big chunks of metal - it is mind-blowing to watch it come together.

This weekend we're running our first big Creativity contest, aimed at visual artists drawing League Of Legends fan-art. If you, or someone you know, has that kind of talent, I hope to see you streaming soon!

Thanks to Emmett Shear, Colin Carrier and Tatiana Moorier for reading drafts of this.