Concept
Outlier opinions
Also: outlier opinion, outliers, minority opinions, dissenting views
An outlier opinion is a view held by only a few people that clearly differs from the majority; in complex situations it is often an early signal, not noise.
Every survey has them: the few answers far away from the rest. Statistics calls them outliers, and the usual move is to remove them so the average looks cleaner. That makes sense when you measure something stable, like the weight of parts coming off a production line. It makes less sense in an organisation, where people and their views keep changing.
Noise or signal?
In a complex system, the outlier is often the first person to notice something. The customer who uses your product in a way nobody planned. The engineer who says the new process won't hold. Most outlier opinions turn out to be nothing. A few are where the next problem, or the next opportunity, starts. You can't tell in advance which ones, so you need to see them all.
Why outlier opinions stay hidden
They rarely make it to the top. Averages smooth them out. Hierarchies filter them, because nobody likes to carry bad news upwards. And people hold back an opinion they know is unpopular, especially when their name is attached to it.
How to hear them
Collect many short, anonymous stories instead of a few interviews, and let people interpret their own story, for example with SenseMaker. Outlier opinions then show up as points away from the clusters, each with the story behind it. When several outliers point the same way, that's a weak signal worth a closer look.