Flight Risk

Why Flight Risk Matters

The average departure is not what hurts. Losing a specific person from a critical role, or a cluster of people who all hold a scarce skill, is what stalls a project or a strategy, and the headline attrition number never sees it coming. Flight risk is the attempt to see it coming, one person at a time.

Consider an engineer whose niche skill has suddenly gone into heavy demand, whose pay has quietly drifted below what that skill now commands, and who is a few years into the role, the point where people tend to look around. No single one of those facts is alarming. Together they describe someone the market is actively pulling at while her own company underpays her. Read it early and you can have a conversation, or fix the pay gap, while she is still in the building.

The mistake worth naming is treating the resignation as the first data point. By then the decision is made and the leverage is gone. Flight risk only earns its keep if it changes what you do before the person decides, which means feeding it into retention and workforce planning rather than into an exit interview.

How Flight Risk Works

Flight risk estimates the likelihood a specific person or segment leaves, by combining signals that each carry part of the story. Tenure and its inflection points matter, since risk clusters around anniversaries and just after them. Market demand for the person's skills matters, because a scarce, highly marketable skill means more open doors. And engagement signals, the softer read on whether someone is still invested, round it out. No single signal is decisive; the estimate comes from how they stack.

The line to hold is that flight risk is a planning input, not a verdict on an individual, and treating it as the latter causes real harm. Acting on a prediction as if it were fact, quietly sidelining someone the model flags, can create the very departure it forecast, and it raises obvious fairness and privacy questions. A worked use: a model flags rising risk among a skill group whose market pay has jumped, so the response is a targeted retention and pay review for the group, not a decision about any one named person. The value is in the segment-level signal that buys time to act.

How to Spot Flight Risk Early

Aim it where a loss does the most damage: critical roles and scarce skills, not the whole population at once. Benchmark each of those roles' pay and market demand against the outside world, then layer in the internal signals of tenure, mobility, and engagement.

The point is to act on the combination early, through a development move, a role change, or a market pay correction, rather than waiting for certainty. Flight risk is a probability, not a verdict, and the whole value is in responding while the person is still deciding.