customer-interview-synthesislisted

# Customer Interview Synthesis Ten interviews produce roughly a hundred quotable moments and one or two findings. The distance between those two numbers is where the work is, and it is where most synthesis fails. The default failure is not laziness; it is that the person doing the synthesis already had a view before the first interview, and the corpus is large enough and varied enough to supply supporting quotes for almost any view someone brings to it. The output looks like evidence, reads like evidence, and is a prior belief with citations. The second failure is the loudest interviewee. One articulate customer who described a problem vividly, at length, and with a memorable phrase will dominate a readout out of all proportion to how many people share the problem, because vividness is what survives in the analyst's memory between reading and writing. Whole roadmaps have been shaped by a single well-spoken participant nobody counted. Both failures produce the same downstream cost. A team builds the wrong thing for two quarters, and because the decision was made on research it is harder to reverse than one made on instinct, since reversing it means admitting the research was wrong rather than that a guess was wrong. The method here is mechanical on purpose: counting, coding, and keeping the verbatim attached, so that the analyst's memory is not the instrument. ## When to use this, and when not to Use it whenever there is qualitative input from more than two sources that h