Justin Kan strapped a webcam to a baseball cap in March 2007 and broadcast his life to the internet. Walking around San Francisco. Working at his laptop. Sleeping. He was 23 and had just come through Y Combinator with a failed web calendar app (Kiko, killed by Google Calendar, sold on eBay for $258,000). The webcam thing was the next bet.
It got attention. Ann Curry interviewed him on the Today Show. Nightline covered it. People compared it to The Truman Show. For a four-person startup with no revenue and no product beyond a guy wearing a camera, the press was enormous.
Then viewers started calling the police on him. Someone spoofed his phone number and reported a stabbing at his apartment. Cops showed up with guns drawn. Kan was at his laptop, hip-hop playing, clothes on the floor. “Did somebody get stabbed in the chest in here?” The next day, someone called in a fire. SFPD put his number on a confirmation list. This was March 2007. The word “swatting” didn’t exist yet. It arguably started in that apartment.
Eight months in, the audience wasn’t growing. People would tune in, watch Kan go to a coffee shop, get bored, leave. So the team opened the platform. Anyone could broadcast. Channels popped up across categories. By 2008, a million registered users. And one category kept pulling away from the rest: gaming.
In June 2011, they spun the gaming section off as TwitchTV. In August 2014, they killed Justin.tv entirely to focus on Twitch. Amazon bought it for $970M that same month, outbidding Google’s $1B offer.
That’s the story everyone knows. Lifecasting failed, gaming won, Amazon got the deal. But something happened recently that makes the story more complicated.
Sponsored by HubSpot
100+ ChatGPT Prompts to Revolutionize Your Day
Supercharge your productivity with HubSpot’s comprehensive guide. This free resource is your fast track to AI mastery:
- Industry-Specific Use Cases: 15+ real-world applications across various sectors
- Productivity Guide: 21 best practices to 10x your efficiency with AI
- Prompt Powerhouse: 100+ ready-to-use prompts for immediate implementation
- Challenge Buster: Overcome common AI hurdles with expert strategies
Plus, in-depth sections on email composition, content creation, customer support, and data analysis.
“Just Chatting” is now the biggest category on Twitch. Not League of Legends. Not Fortnite. Not GTA. Just Chatting. People sitting in front of cameras talking to their audience. 2.9 billion hours watched over the past year. Nearly triple the next gaming title. IRL (In Real Life) content, which is literally people broadcasting their daily lives, grew 104.9% year over year.
Lifecasting is the number one thing on the platform that was built because lifecasting failed.
That’s worth sitting with for a second. Kan’s original concept, streaming your daily life to strangers on the internet, turned out to be exactly right. He was just early. The technology (smartphones with front-facing cameras, mobile broadband, established streaming infrastructure) didn’t exist in 2007. The behavioral norm (watching random people live on the internet) hadn’t been established. By the time Instagram Live launched in 2016 and TikTok started eating attention in 2018, the audience was ready. And Twitch, the platform that won by pivoting away from lifecasting, has now pivoted back.
The iPhone launched the same year as Justin.tv. Without a front-facing camera. The selfie camera came in 2010 with the iPhone 4. So when Kan was taping a webcam to a baseball cap and carrying a laptop in a backpack, the entire infrastructure for casual mobile broadcasting was still three years from existing at the hardware level. Kyle Vogt, his co-founder, hand-built the camera rig. That rig was doing what an iPhone would do natively within a few years. Kan wasn’t wrong about human behavior. He was wrong about the timeline.
But the gaming detour is worth understanding on its own terms, because the mechanics of why gaming works as spectator content explain something broader about what makes any content watchable.
On paper, watching someone else play a video game makes no sense. You’re not playing. You have no control over the outcome. The graphics look the same whether you’re watching or playing. Why wouldn’t you just play the game yourself?
The answer is the same reason people watch sports instead of playing them. Spectating works when there’s skilled performance combined with uncertain outcomes and a personality delivering it. A StarCraft player executing 200 actions per minute is doing something most viewers literally can’t do. A speedrunner completing a game in 45 minutes that took the audience 40 hours is demonstrating mastery. The gap between the viewer’s ability and the performer’s ability creates the entertainment value.
Lifecasting in 2007 didn’t have this. Watching Kan walk to a coffee shop involved no skill, no stakes, and no uncertainty. There’s nothing about daily errands that creates the gap between performer and audience that makes spectating compelling.
Gaming had all of it. Skill (high-level play), stakes (competitive matches, speedrun attempts), uncertainty (will they win? will they die?), and personality (the streamer talking, reacting, raging, celebrating in real time). Add interactive chat where the audience can talk back and influence the streamer, and you have something closer to a new form of live entertainment than to traditional video.
And yet “Just Chatting” overtook gaming. So the skill/stakes/uncertainty framework doesn’t fully explain it. Or rather, the framework evolved. The streamers who dominate Just Chatting, people like Kai Cenat ($52M in estimated 2025 earnings), aren’t broadcasting mundane daily life. They’re performing. Cenat runs month-long “Mafiathon” events with celebrity guests, orchestrated chaos, and the production values of a talk show. The “chatting” is the format. The content is entertainment.
The difference between Kan’s lifecasting and Cenat’s Just Chatting is the same difference between someone recording themselves at dinner and a cooking show. Same format. Completely different product. The format (live + personal + interactive) was always right. The execution in 2007 was too raw, too unstructured, and too early for the infrastructure and audience to support it.
The founding team’s trajectory after Justin.tv tells you something about the density of talent in early-stage startups that most people underestimate.
Four co-founders. Justin Kan went on to start Exec (on-demand personal assistants), then Atrium (AI law firm, shut down 2020), then Fractal/Stash (gaming). Emmett Shear ran Twitch as CEO for 12 years, then served as OpenAI’s CEO for five days during the Sam Altman firing saga in November 2023 before Altman came back. Michael Seibel became CEO of Y Combinator. Kyle Vogt, the guy who wired the webcam to the baseball cap, co-founded Cruise Automation, which GM acquired for over $1 billion.
The camera rig guy built self-driving cars. The CEO ran OpenAI for a week. The other co-founder ran the most influential startup accelerator on earth. Four people from a lifecasting experiment, and three of them ended up at the helm of billion-dollar operations or institutions. Paul Graham, who funded them, said: “This is very common in the history of startups: You get 9 out of 10 things right.”
The implication is that the team matters more than the idea. Justin.tv’s idea was wrong (or early, depending on your read). The team was right. And the team’s ability to notice that gaming was growing, spin it out, kill the original product, and commit fully to the spinoff is the skill that turned a failing lifecast into a $970M acquisition. Kan called it “emergent strategy”: you don’t plan where to go, you watch where the traffic flows and follow it.
Twitch’s numbers today: 240 million monthly active users. $1.9B in revenue for 2025. 2.05 million average concurrent viewers. 7 million monthly streamers, though only about 920,000 of them earn any money at all. The top 1% takes more than half of all payouts.
That last number matters. Twitch created the “middle class of gaming,” where someone with 500 concurrent viewers can earn $3,000-5,000/month from subscriptions, ads, and donations. That career path didn’t exist before 2011. But the middle class is smaller than the narrative suggests. 72.6% of monthly streamers earn zero. The economic structure looks more like Hollywood (a few stars, a thin middle, and a massive base working for free) than like a democratized creator economy.
And Twitch’s position is eroding. Market share dropped from 70% to 54% as Kick (backed by gambling company Stake, offering a 95/5 revenue split to creators) grew 131% in 2025, reaching 4.5 billion hours watched. YouTube Gaming grew 12% to 8.8 billion. The 50/50 subscription split that built Twitch’s creator ecosystem is now its competitive vulnerability. When Kick offers 95/5 and YouTube offers 70/30 from day one, the $2,240 monthly difference per 1,000 subscribers between Twitch and Kick is real money that drives migration.
Amazon used Twitch Prime (free channel subscription bundled with Amazon Prime) to connect its commerce engine to the youngest, most engaged audience on the internet. The acquisition wasn’t about Twitch’s revenue. It was about reaching a demographic that doesn’t care about free shipping on paper towels but does care about supporting their favorite streamer. Whether that strategic value holds as Twitch’s share shrinks is an open question Amazon is probably thinking about right now.
The circular irony of Justin.tv is hard to overstate. Kan invented lifecasting. It failed. The platform pivoted to gaming. Gaming built a $970M acquisition and a decade of dominance. And now the biggest category on the platform is… people sitting in front of cameras talking about their lives.
If Kan had started Justin.tv in 2016 instead of 2007, with an iPhone instead of a webcam taped to a hat, with an audience already trained by Instagram and Snapchat to watch strangers live, he might not have needed the gaming pivot at all. The product might have worked as designed. The creator economy around “Just Chatting” and IRL content might have emerged directly from the lifecasting concept without the gaming detour.
But then there wouldn’t be Twitch. The gaming detour built the infrastructure (streaming technology, chat systems, creator monetization, moderation tools) that makes Just Chatting work today. The failure created the platform that eventually made the original idea viable.
Sometimes being wrong is the only way to build the thing that makes you right later.
Further reading
- The Many Pivots Of Justin.tv (Fast Company, 2013): The definitive early account. Swatting incident. Paul Graham quotes. The pivot from lifecasting to gaming. Written a year before the Amazon acquisition.
- Twitch Statistics 2026 (SociallyIn, March 2026): Current numbers. 240M MAU. Just Chatting as #1 category. IRL growing 104.9%. Market share decline from 70% to 54%. Kick’s 131% growth.
- Justin.tv (Wikipedia): The founding details. Kyle Vogt’s camera rig. October 2007 open platform. Gaming section dominance. Shutdown in August 2014.
That’s all for today. I’m going to go stare at my laptop and pretend I’m thinking strategically. See you Monday.
Leave a Reply