Why we picked it This is the canonical talk on why early founders should do unscalable, manual work to get and understand their first users, told by people who lived it (DoorDash's Stanley Tang, Teespring's Walker Williams, Twitch's Justin Kan). Tang describes running DoorDash by literally testing demand with a landing page and then delivering orders by hand, which is a concierge MVP in practice. It is the honest counter to the instinct to build automation before you know anyone wants the thing.
Doing Things That Don't Scale, PR, How to Get Started (How to Start a Startup, Lecture 8)
On Y Combinator / Stanford CS183B by Stanley Tang, Walker Williams, Justin Kan ~50 min
- Test demand first with something tiny (a landing page, a few calls) before building, then serve early users manually even if it does not scale.
- Doing things by hand is how you learn what customers actually need, not a failure to be embarrassed about.
- You find a way to scale later, the early job is delighting a small number of real users.