Weak signal detection

A weak signal is a hint that something is starting to shift. A handful of customers describe the same workaround. A few people in one team stop raising problems. A complaint sounds different from the usual ones. On its own, each of these looks like noise. By the time it shows up on a dashboard, it has usually turned into a strong signal, and into a problem.

Why averages hide weak signals

Surveys and KPIs summarise. They are built to show the middle, so the few answers at the edge disappear into the average. In a complex system, that edge is often exactly where change begins. The strategy researcher Igor Ansoff made this point in 1975: organisations that wait until a signal is certain respond too late.

How stories make weak signals visible

Weak signals rarely arrive as numbers. They arrive as small stories, micro-narratives, about something people noticed. When many people share such moments and interpret them themselves, for example with SenseMaker, patterns appear: a cluster that keeps growing, an outlier opinion that keeps coming back, a non-dominant narrative that contradicts the official one. Because the storytellers interpret their own stories, the signal isn't filtered through someone else's assumptions first.

What to do with a weak signal

A weak signal is not a forecast. It is a reason to look closer. In complex situations the useful response is small: a safe-to-fail experiment that tests whether the pattern is real, so you can amplify it or dampen it depending on what happens. The Weak Signals Radar is an ongoing story collection built for exactly this.