A resignation is rarely a sudden decision. Between the first doubt and the letter, several months usually pass. During that period the person keeps working, but their relationship with the organisation changes: fewer initiatives, less disagreement voiced, a shorter horizon. By the time the decision shows up in HR indicators, it has already been made.
Definitions and calculation
The turnover rate relates departures to average headcount over a period, usually a year: number of departures divided by average headcount, multiplied by one hundred. The raw figure is of limited use. What informs a decision is its segmentation:
- voluntary turnover (resignations) versus involuntary (dismissals, contract ends)
- avoidable versus unavoidable departures (retirement, relocation)
- regretted versus non-regretted: losing a key profile does not weigh like an expected exit
This article focuses on the segment leadership can act on: avoidable voluntary turnover, concentrated on the people the organisation wanted to keep.
What a departure really costs
Published estimates vary widely by role, from a few months to more than a year of salary for expert or managerial profiles. Rather than a generic figure, the useful calculation uses your own components:
- direct costs: recruitment, onboarding, training the replacement
- productivity loss: ramp-up time, vacancy period
- workload shifted onto the remaining team, at the risk of wearing out those who stay
- loss of undocumented knowledge and of client or supplier relationships
Added to this is an effect documented by research on turnover contagion: when a credible person leaves, leaving becomes a legitimate question for those who were hesitating. Clustered departures within one team are its classic symptom.
Why classic indicators look backwards
The turnover rate is an outcome indicator: it counts decisions already made. The annual review captures prepared speech, in a setting where voicing doubt carries a cost. The engagement survey photographs individual perceptions once a year. None of these tools observes the period during which the decision is built.
Yet that period leaves traces at the collective level: gradual withdrawal of voluntary contributions, less disagreement expressed in meetings, a growing gap between what the team understands of the strategy and what it is asked to do, workload settling in without arbitration. We detail the nature of these signals in our article on weak signals; this one is about turning them into retention steering.
Steering retention as an organisational matter
As long as turnover is treated as a sum of individual cases, the response stays individual: counter-offer, defensive promotion, retention interview. These responses come after the decision and do not address what produced it. An accepted counter-offer rarely retains beyond a few months when the cause of the departure has not moved.
Steering retention at organisational level involves three shifts:
- measuring collective dynamics continuously, team by team, rather than an annual company-wide average
- qualifying each finding with the team concerned before acting: a signal is not a diagnosis
- deciding little, but completely: one action at a time, an accountable person, a deadline, a verification criterion
One framing point is essential: this measurement covers collective aggregates, never individual behaviour. Headcount thresholds limit re-identification risk in small teams, the setup is presented to employee representatives, and it does not predict any individual's departure: it makes the state of a collective visible. The distinction is not only ethical; it conditions the reliability of the measurement, because a setup perceived as surveillance produces façade data.
Where to start
Start with a team where the stakes are real: high value, a history of departures, dependence on a few key people. Measure the collective dynamic, qualify the findings with the team, choose one action and verify its effect. Retention is won in day-to-day functioning, not in exit interviews.
To frame this work for your context, we offer a 30-minute discussion from which you leave with a structured reading of your turnover: relevant segments, priority teams, and a first measurement scope.