The Danger of Averages in People Data
Averages are comforting. They simplify complexity, reduce noise and offer leaders a sense of control. In people data, they are often treated as the safest way to understand how things are going overall. A single score can be shared easily, tracked over time and discussed without too much friction. It creates the impression that reality has been captured.
The problem is that averages are designed to smooth variation and variation is where most people issues live.
Averages don’t show where pressure is building. They show where it has already evened out.
In organisations, emotional experience is rarely evenly distributed. One team can be thriving while another is quietly struggling. A small group can be under intense pressure while the majority feels broadly fine. When these experiences are rolled up into a single number, the result often looks reassuring even when something important is happening beneath the surface.
This is why leaders can review people data, feel reasonably confident and still be surprised later by outcomes. The data was not wrong. It was incomplete in a very specific way. It described the middle and obscured the edges.
People data is particularly vulnerable to this because it is often collected infrequently and analysed retrospectively. By the time responses are aggregated, any early signals have already been diluted. What remains is a stable-looking average that hides movement rather than revealing it.
Stability in the data does not always mean stability in reality.
Another danger of averages is that they flatten experience across time. A score taken at one moment can hide rapid change underneath it. A team moving quickly from coping to struggling may still report a similar overall score, even though the direction of travel has shifted significantly.
Leaders reading that score are left with an incomplete picture. They see where things were, not where they are heading. Decisions are then made based on reassurance rather than anticipation.
This creates a subtle but important leadership risk. When averages dominate attention, intervention is delayed until issues become widespread enough to affect the overall number. At that point, the opportunity for light-touch response has often passed. What could have been addressed through conversation or small adjustment now requires more visible action.
By the time an average moves, the issue has usually been present for a while.
There is also a behavioural effect. When leaders focus on averages, teams learn what gets noticed. Issues that affect only a few people are less likely to be raised if they do not move the headline number. Over time, this discourages early escalation and reinforces a culture of coping rather than addressing.
This is rarely intentional. It is a consequence of what the system rewards.
None of this means that averages are useless. They have a role in providing orientation and context. The danger comes when they are treated as the primary signal rather than a starting point for deeper understanding.
Managing people well requires attention to variation, not just central tendency. It requires leaders to be curious about who sits above and below the average, where experiences differ and how those differences are changing over time.
What matters most is often happening in the minority, not the majority.
When leaders supplement averages with earlier, more granular insight, their behaviour changes. They ask different questions. They notice patterns before they harden. They are able to respond in ways that feel proportionate rather than corrective.
The danger of averages in people data is not that they are wrong. It is that they create a false sense of completeness. They make leadership feel informed while leaving important parts of reality unseen.
Recognising this shifts the role of people data. Instead of asking whether the score is up or down, leaders begin to ask where pressure is emerging and who is experiencing it first. That shift is where better people management begins.
This essay in context
This essay explores why averages often hide the very issues leaders most need to see.
The wider series examines how noise distorts attention, how emotional strain becomes operational risk, and why managing people requires different signals than managing performance.
Together, they describe what changes when organisations stop relying on averages as a proxy for understanding and start treating variation as a source of insight.
