The brief says users are dropping off in onboarding. It doesn’t say how anyone knows that.
Someone hands you a brief: redesign the onboarding flow, users are dropping off. Most designers open Figma. I’ve learned to wait about thirty seconds and ask how we know onboarding is the problem.
Briefs arrive looking like answers, sometimes with wireframes already sketched by whoever wrote them, which is always a fun surprise. Underneath there’s usually a pile of assumptions nobody has tested.
The drop-off might not be an onboarding problem at all. Maybe the wrong people are signing up in the first place. Maybe it’s the product, and onboarding is just where people notice. A beautiful new onboarding flow for the wrong cause just gets you to the same drop-off faster.
So treat the brief as a hypothesis and poke at it a little. None of this needs a three-month research phase every time someone asks for a new filter. You just find the assumption the whole thing is resting on. There’s always one. Usually it’s that the problem is where everyone thinks it is, or that fixing it will move the number we care about.
Then decide how expensive it would be to be wrong. On small bets, go ahead and adjust as you learn. On big ones, where being wrong means several engineers building the wrong thing for a quarter, spend a week finding out first.
References