Twitch, YC Founder Story

Company: Twitch (spun off from Justin.tv) Founders: Justin Kan, Emmett Shear, Michael Seibel, Kyle Vogt YC Batches: Summer 2005 (Kiko) → Winter 2007 (Justin.tv → Twitch) Lead Founder for Twitch: Emmett Shear (CEO) & Justin Kan (Co-founder) Industry: Live Streaming / Gaming / Creator Economy Twitch Launched: June 6, 2011 Acquired by Amazon: August 2014 for $970 Million Peak Stats: 55 million unique visitors/month, 4th largest US internet traffic source

---

---

The One-Line Summary

Justin Kan and Emmett Shear failed twice, sold their second company for pennies on eBay, strapped a webcam to Justin's head to prove live streaming was possible, and then watched their gaming vertical quietly overtake everything they had built, becoming a $970 million company that nobody thought was a real business.

---

Lens 1, The Before State

Who Were They Before YC?

Justin Kan and Emmett Shear were Yale roommates who bonded over two things: video games and the idea of building technology companies. They were not engineers from Stanford with Silicon Valley connections. They were East Coast university kids who had to convince Paul Graham to even look at them.

After graduating from Yale in 2005, Justin, who had grown up in Seattle watching his parents run their own business, with an early dream of becoming a CEO, made his first move into startups with Emmett. Their idea was a web-based calendar app called Kiko, designed to integrate with Gmail and other email clients. It was a smart idea, practically executed. They got into YC's very first batch in Summer 2005, which also included Reddit and Loopt. YC invested $12,000.

Kiko launched in August 2005, raised an additional $50,000 from investors, and showed real promise. Then Google released Google Calendar for free. Within months, Kiko was irrelevant.

The Personal Pain They Were Living

There was no dramatic personal pain story behind Kiko. It was a good idea that got crushed by a tech giant's free product, a brutal but common startup death. What mattered about Kiko was not what it built, but what it taught: that timing, market awareness, and the willingness to kill a failing idea fast are as important as the idea itself.

Faced with a dead product and investor money they needed to return, Justin and Emmett did something nobody in the startup world had done before: they listed Kiko on eBay. For the first few days of the auction, nothing happened. Then, in the final hours, bids rolled in, and they sold the company for $258,000, enough to repay their investors and walk away with dignity. Kiko became only the second company ever sold on eBay at the time.

Paul Graham, watching this unfold, was so impressed with how they handled the failure that he asked them: "Do you have any other ideas?"

Coincidentally, they did.

Why They Almost Didn't Look Like YC Material

By the time they came back to YC with Justin.tv in 2007, the story didn't make obvious sense. The pitch was essentially: "Justin is going to wear a webcam on his head and stream his entire life online, 24 hours a day, 7 days a week."

This was not a technology pitch. It was a publicity stunt masquerading as a startup idea. The product, lifecasting, had no clear business model, no precedent, and no guarantee that anyone would watch. What it had was audacity. And in that audacity, it had a deeper technological thesis: that live streaming over the internet, at consumer scale, was about to become possible, and whoever built the infrastructure first would own the space.

Paul Graham funded it anyway.

Key Insight for Aspiring Founders

Failing well is a superpower. Justin and Emmett didn't just fail at Kiko, they failed with so much creativity and composure (selling on eBay, repaying investors) that their investor came back and asked for another shot. How you exit a failure tells funders everything about how you'll handle the next one.

---

Lens 2, The Idea Origin

How Justin.tv Was Actually Born

After the Kiko sale, Justin and Emmett were watching the rise of YouTube and asking the same question everyone was asking: what's next after video on demand? Their answer: live video. Real-time, unscripted, continuous. Not highlights, the actual thing happening now.

The experiment they designed was deliberately extreme. Justin would strap a wireless webcam to a baseball cap, connect it to a laptop in a backpack running on four EVDO data cards, and stream his entire life uninterrupted from March 19, 2007 onward. He would only remove the camera for the bathroom and sleep. Everything else, meals, conversations, parties, arguments, getting pulled over, police raids triggered by false emergency calls from viewers, was live, on camera, on the internet.

This was technically insane. The internet infrastructure of 2007 barely supported it. Streams dropped constantly. Video quality was terrible. But the concept worked: people watched. And more importantly, the team had proved to themselves that the infrastructure problem was solvable.

They joined YC's Winter 2007 batch with this concept, this time with two additional co-founders: Michael Seibel, who handled business and operations, and Kyle Vogt, who later went on to found Cruise Automation (acquired by GM for over $1 billion).

The Gaming Category Nobody Asked For

Justin.tv grew into a general live-streaming platform where anyone could create their own channel and broadcast whatever they wanted. By March 2008, the team added category pages, sports, news, entertainment, gaming. Nobody expected gaming to be anything. It was just one of many tabs.

Then StarCraft 2 was released. And League of Legends. And a small but intensely engaged community of gamers started using Justin.tv not to play, but to watch other people play. The numbers in that one category started growing faster than everything else combined. By 2010, gaming was unmistakably the dominant vertical on the platform, pulling traffic no other category came close to matching.

The idea for Twitch wasn't invented, it was observed. The users showed the founders what they wanted before anyone thought to ask.

The First Ugly Version of Twitch

In January 2011, facing pressure to monetise Justin.tv more effectively, the team launched two internal "skunkworks" projects: Justin.tv Mobile (led by Michael Seibel) and Justin.tv Gaming (led by Emmett Shear). The deal they made among themselves: after six months, whichever project had the most traction would determine the company's future.

Justin.tv Gaming won, decisively. On June 6, 2011, it spun out as TwitchTV, named after the term "twitch gameplay", the fast, reflex-based play style of competitive gaming. Within two years, it had 35 million unique monthly visitors.

The Pattern This Follows

Twitch fits YC's most underrated idea archetype: let your users show you the pivot. The founders didn't decide gaming was a billion-dollar opportunity from data. They noticed a behaviour pattern, people watching gameplay, that existed before a dedicated product did. The job was to follow the signal, not manufacture one.

---

Lens 3, The Application Anatomy

Two YC Applications, What Changed

First Application (Kiko, Summer 2005): Justin and Emmett applied to YC's inaugural batch with a calendar app. The pitch was straightforward, a better online calendar that integrated with email. They had a working product, a clear use case, and a credible team. Paul Graham funded 8 companies from roughly 225 applications that summer. Kiko was one of them.