2024-09-30
Sampling bias in win-loss interviews (and how to say it out loud)
Customer Listening Loops graduates often uncover uncomfortable skews. Mara explains how to disclose bias without eroding trust.
By Mara Ellison
Every interview panel inherits skew: timing, incentive design, or who agrees to talk. Name the skew in the appendix of your synthesis memo before leadership invents a rosier story.
Pair qualitative findings with the sample size sanity checklist from Advanced Cohort Diagnostics. Even non-statisticians can understand when a story is built on thin air.
Offer leadership two decisions: proceed with partial insight or fund another wave of interviews. Framing the trade-off respects their time.
When bias cannot be fixed quickly, document the mitigation you will use until better data arrives — usually a conservative forecast and a visible research backlog.
Tagged: research, ethics