Skip to content

Flight risk

Flight risk is the likelihood that a particular person leaves within a period that matters to a plan. It is most often treated as an attribute of the individual and is better read as a relationship between what the role now offers and what that person is looking for.

Flight risk is usually created by a change in the role rather than a change in the person.

Why it matters when the plan changes

Departures cluster around the events that make a plan fragile: a change of ownership, a new operating model, a role whose content shifted. Those events change what a job contains without changing its title, and people leave because the work stopped meeting the internal disposition toward work that drew them to it in the first place, which looks from outside like an unrelated personal decision. Risk peaks precisely where a plan demands behaviour the role no longer supports, not where a person has simply changed.

The tension is that flagging individuals as flight risks is both useful and hazardous. It invites treating a person differently on the basis of a prediction they never saw, and in several jurisdictions it raises questions under data protection law. The defensible version describes exposure at the level of a role and a dependency, not a watchlist of names, and measures it the way turnover itself is measured: against a defined period, not a verdict on a person.

In practice

An operating-model change moves a leader from running her own unit to coordinating across three. She performs well and leaves in eleven months. The exit interview records a better offer. What changed was that the work she was drawn to stopped being in the job, and no part of the change plan had looked at that.

Evidence

What it cannot tell you

Flight risk describes exposure in an arrangement, not a fact about a person's future. It cannot say when someone will leave, only that the conditions for leaving are present. It is silent on causes outside the role itself, such as personal circumstance, and treating it as a prediction about a named individual overstates what it can support.

Questions

Exposure can be reasoned about, not predicted with confidence. Wikipedia's 2026 entry on Work motivation describes it as a person's internal disposition toward work, which a changed role can stop meeting. A role whose content has shifted away from that disposition, under a plan that depends on the person, is a real exposure.

Recording a prediction about whether a named employee will leave is processing personal data for a purpose they were probably not told about, and it may affect how they are treated. It needs a lawful basis, transparency and in many cases a works council conversation.

As a property of the arrangement rather than a person. Turnover, as Wikipedia's 2026 entry on Turnover (employment) frames it, is measured against a defined period; exposure should be expressed the same way, as a role whose content has changed under a plan with no alternative if it becomes vacant.

Changing the role back toward what the person was drawn to, distributing the dependency so their departure is survivable, or having the conversation about what the job has become. The first two are structural; the third is usually the one nobody has had.

It delays rather than resolves, which can be worth buying when the delay covers a critical window. It does not change the reason the person is leaving, so the exposure returns when the payment period ends, usually with less notice than the first time.