You've done the work. Sixteen interviews, all pointing in roughly the same direction. Your pattern is solid. You write it up. You feel good about it.
Then in the debrief, someone asks about P4, P9, and P13 — the three who said something different. You've noted them as outliers. That's the professional thing to do.
But here's what I've learned after thirty years of this: the participants who contradict your pattern are almost never random. They're a signal. Sometimes they're a different mindset that cuts across your cohorts entirely. Sometimes they're the reason your client's brief is wrong. Occasionally they're the most important finding in the project — and they nearly got filed under "exceptions."
The question isn't whether the outliers matter. It's whether you had the infrastructure to find out before you wrote the report.
The problem isn't researcher skill. Senior researchers know this intuitively. The problem is that finding out why three people said the opposite thing requires going back through sixteen transcripts, cross-referencing what else those three said, looking for a pattern in their pattern — all under deadline, all manually.
So we make a judgment call. We note the exceptions. We move on. And sometimes we're right. But sometimes we've just buried the most interesting thing in the data.
What this looks like in practice
A few months ago I was working on a spirits brand study — eighteen interviews across three cohorts. The dominant pattern was clear: discovery is social. Fourteen participants described finding the brand through a recommendation, a moment at a bar, someone ordering next to them.
Three participants were different. They'd found the brand through editorial — a magazine feature, a newsletter, a piece of long-form content. In every other respect they looked like the rest of the sample. Same age, same consumption patterns, same cohort.
I almost filed them as channel outliers and moved on. Instead I looked harder at what else they had in common. They were all what I'd call considered discoverers — people who research categories deliberately before entering them. The social-discovery group was responsive. This group was deliberate.
That distinction changed the media recommendation entirely. Not a small adjustment — a different channel strategy for a meaningful segment of the market. It was sitting in the three people who said the opposite thing.
The discipline this requires
Catching this consistently isn't about being smarter. It's about having a system that makes counter-evidence visible by default — not something you have to go looking for under pressure.
The observation isn't new. The infrastructure to act on it reliably is what's been missing.
Greg Mitchell · Sharpen Consulting Elliott is built to surface the three who said the opposite thing — automatically. Beta access open now for senior qualitative researchers. Join the beta →