Evan Spiegel, Snap's CEO, argues distribution now matters most, since software is easy to copy and network effects alone are an insufficient defence.
Spiegel says Snap's design team is flat and deliberately bottlenecked, with designers owning product work and leaders staying close to the product.
Snap must prove in 2026 that it can be a strong, profitable business, as it is not yet net income profitable despite roughly 200m monthly game players.
Snap's Spiegel says distribution now outweighs every other moat
Lasting consumer social products now depend on distribution, and Snap is betting on ecosystems and hardware to secure it. Software is easy to copy and network effects are not enough, so Snap leans on creators, developers, AR and Specs. Spiegel says 2026 is the year the company must also prove it can be profitable.
Takeaways
Distribution is now the most important moat. Spiegel says founders spend too little time on distribution. Snapchat grew by linking people to their closest friends, TikTok won by spending billions subsidising both sides of its video marketplace, and Threads leaned on Meta's other products.
Network effects alone cannot defend software. Spiegel calls network effects important but insufficient. Snap added creator and developer ecosystems, an AR platform with millions of lenses, and vertically integrated hardware. Snapchat Plus, which he sees Instagram copying, has 25m subscribers and over a billion dollars in run rate.
Specs are Snap's bet on post-phone computing. Spiegel believes phones isolate people. Spectacles moved from a 2014 camera-first device to one with a second camera for depth and a display, plus a 2024 developer operating system. Consumer Specs arrive this year. He dislikes heads-up displays because of where the screen sits.
Flat teams counter big-company risk aversion. Drawing on Safi Bahcall's Loonshots, Spiegel argues large hierarchical organisations make people risk averse, so leaders must bridge them with flat teams. Snap's design team of nine to 12 people is flat and reviews work weekly.
Snap builds AI agents for clearly defined jobs. Snap's automated code review has caught close to 10,000 bugs, a go-to-market agent runs on Claude, and Spiegel uses a Glean-based agent to review dashboards. He expects human adoption to shape AI more than technologists think, with societal pushback.
2026 must prove Snap is profitable. Spiegel calls the year a crucible. Snap is not yet net income profitable, though about 200m people play games on Snapchat each month. He calls Snap the middle child of social, bigger than Pinterest and Reddit but far smaller than Meta and Google.
How Snap runs product and design
Design is a deliberate bottleneck for shipping, and designers were meant to own product work rather than only produce visuals.
Each new designer presents a portfolio piece on their first day, explaining why they made it and what they learned.
Snap waited to hire product managers and now uses them for coordination at scale.
Users asked for a send-all button, so Snap built Stories with no public metrics, a 24-hour lifespan and chronological order.
Tap a name to go there · tap a line to see why they are linked
On this episode 2
Show 1
Spiegel favours hour-long conversations about how people use technology over surveys, and rejects Keith Rabois's advice not to talk to customers.