Distribution, not software, is the real moat — Snap defends it with ecosystems, hardware and a tiny design team.
Evan says distribution is the hardest thing to figure out. When Snapchat launched, the App Store was new and people downloaded and tried new apps all the time; that is not the case now. Consumer teams obsess over product market fit and spend nearly no time on distribution, which he calls a huge differentiator.
Back then people believed network effects meant the biggest networks could not be beaten. Snapchat found that what really mattered was connecting you to the right people, not to everyone: your best friend, your partner, your spouse. That, plus mobile and App Store distribution, is what allowed it to grow early on.
TikTok did it with money, which Evan found innovative: billions of dollars subsidizing both sides of its video marketplace, acquiring customers to watch videos and paying creators to make them, bootstrapping the ecosystem. Threads leveraged the distribution Meta already had across its other products. Both are distribution stories, not only product stories.
Fifteen years ago Snap learned that competitors cloned their software features easily, something everyone is discovering today with AI. So they built ecosystems instead: relationships between creators and Snapchatters, and an AR platform where developers built millions of lenses. They also invested in hardware, where a vertically integrated stack is very hard to copy.
Customers kept asking for a send-all button, and separately said social media felt permanent and judged through likes and comments. People also found reverse chronological feeds strange, with the end of the birthday party appearing first. Snap built Stories instead: chronological, no public metrics, disappearing after 24 hours so each day starts fresh.
Evan points to Safi Bakal's book Loonshots. Delivering at scale needs a large, hierarchical organization, which makes people promotion-focused and risk averse; innovation needs very flat structures. Great companies run both, and leaders must create a healthy dialogue between them. At Snap that is engineers serving nearly a billion customers and a nine-to-twelve-person design team.
The velocity of design work. Evan meets the designers a couple of hours every week just to look at new work, hundreds of ideas weekly. He combines Stanford product design's empathy with art school's grinding expectation of constant making and brutal critique. Their favorite saying: if you want a good idea, you have to have lots of ideas.
Asked about Keith Rabois's view that you should not, Evan says you have to talk to customers and share your idea as quickly and as frequently as possible. That does not mean taking their advice. He dislikes the survey model; going deep for an hour or two about how technology fits someone's life teaches far more.
Everyone told them disappearing photos made no sense because you can always screenshot. Evan and Bobby realized that if you pressed and held to open a snap and took a screenshot, the phone reported that your finger had lost contact with the screen. They used that touch event to notify the sender, and it spread.
Snap is nearly at Fortune 500 scale and almost a billion monthly active users, and is about to launch Specs after twelve years of investment, yet it is still not net income profitable because of that investing. This year has to prove Snapchat is a strong, growing, profitable foundation, with 200 million monthly game players.
“People don't spend nearly enough time thinking about distribution and figuring out distribution.”
“So it's very easy to copy software features, it's very hard to copy or to replicate a full ecosystem or a platform.”
“you know if you want to have a good idea, you have to have lots of ideas.”
“humanity dictates how technology is adopted, right?”
“Distribution Over Software.” https://www.youtube.com/watch?v=-7Yol5vX5xw. Transcript summary by WhipScribe, https://whipscribe.com/transcript-pages/distribution-over-software.html.