A handful of conversations can teach you anything
Founders often draw confident conclusions from very small numbers.
Five users said the onboarding was confusing — “our onboarding is broken.” Three customers asked for the same feature — “users want this feature.” Two pre-sales — “there’s clear demand.” One enthusiastic interview — “we’re on to something.”
Sometimes those conclusions are right. Often they’re premature. Small samples carry an enormous amount of variation. Whatever pattern you happen to see in your first few users can look very different by the time you’ve seen many more — and decisions made in the meantime have to be unwound.
This topic is about how to read small samples without being fooled by them.
Why small samples mislead
Three reasons, all of them familiar.
First, rando...
