Testing in waves

Five to eight people per wave. Enough to surface the major issues and see a pattern; few enough that you can change something before the next round.

Talk to too few and you will mistake one person's opinion for a signal. Talk to too many without changing anything and you hit diminishing returns, hearing the same feedback while learning nothing new.

Revise between waves. A research programme that runs twenty identical sessions has run one session twenty times.

No questions on this lesson yet. Highlight a passage to ask about it, or use Ask a question.