How to spend 45 minutes in a whiteboard challenge
Almost nobody fails a whiteboard challenge for running out of time. They fail for spending the time in the wrong proportions — twenty minutes drawing screens and four minutes understanding the problem, when the scorecard weights those the other way round.
The split below is what strong candidates converge on. It is not a rule, and no interviewer is holding a stopwatch. It is a default that keeps you out of the two failure modes that actually cost offers: designing before you have framed the problem, and running dry before you have named a single trade-off.
Minutes 0–5 · Clarify and constrain
Ask until you know what you are solving. Platform, market, which side of the marketplace, what is out of scope. Write the answers on the board — you will point at them later when you justify a decision. Candidates who skip this are optimising a problem nobody set them.
Minutes 5–10 · Pick a user and a moment
Commit to one person and say why. “Remote workers and families and students” means designing for nobody, and every decision after it becomes unfalsifiable. The scored part is not which user you pick — it is that you picked one and can defend it.
Minutes 10–20 · Map the journey, mark the breaks
Draw the current experience end to end and put an X where it fails your user. This is the highest-value block on the board and the one most candidates skip in their hurry to start designing. Before you move on, say out loud which single break is the dominant one. That sentence is what the rest of your answer hangs from.
Minutes 20–30 · Diverge, then converge
Three directions, about a minute each, then pick one and say plainly why you are killing the other two. Interviewers are watching for whether you can generate options at all, and then whether you can choose between them under pressure. Generating three and quietly drifting into the first is worse than generating one.
Minutes 30–45 · Sketch, trade off, measure
Draw the thing. Rough boxes are fine — nobody is scoring your linework. Then do the two things candidates most often run out of time for: volunteer what your design costs and who it hurts, and say how you would know it worked. A primary metric and a counter-metric. Not “engagement” — something falsifiable.
What to do when the clock beats you
Stop and summarise. Say what you would have done next and what you would want to validate first. Interviewers expect unfinished work — a whiteboard challenge is a thinking sample, not a deliverable. Finishing badly reads far worse than stopping cleanly.
Ready to run one properly? Every challenge in the library comes with what it tests, how it is scored, and where candidates lose it.