Interview- and Workshop-Based Data Gathering Playbook

There's no universal number, but a useful rule of thumb is to keep interviewing within a stakeholder segment until you stop hearing new themes — typically 12-20 interviews across a mixed sample surfaces the major themes for most organizational or strategic questions, with diminishing returns after that unless you're deliberately expanding into a new segment.

Take notes as close to verbatim as possible during the session, especially for direct quotes that might later illustrate a finding, but code and synthesize them into summarized themes afterward. Verbatim notes protect you from misremembering nuance; summarized coding is what makes cross-interview patterns visible.

Use structured formats — silent brainwriting before discussion, anonymous dot-voting, and small breakout groups reporting back — rather than open floor discussion, since these formats surface every participant's input before social dynamics can suppress it. If one person still dominates verbally, the facilitator should explicitly invite quieter participants by name rather than waiting for volunteers.

A theme is a recurring topic that shows up across coded input; it becomes a finding once you've checked corroboration — how many independent sources raised it, and whether any contradicted it — and can state it with a confidence level. Presenting a single articulate quote as a finding without checking corroboration is the most common synthesis mistake.

Yes, the same sampling, guide, and coding discipline applies — the main adjustment is that the sampling plan should be built around customer segments (size, tenure, usage pattern) rather than internal seniority and function, and incentives or scheduling logistics for external participants typically need more lead time.