Reading Turnover Properly

Turnover is the most-quoted wellbeing-adjacent metric and the most poorly analysed. The headline rate is close to useless.

The segmentation that matters

Voluntary against involuntary. Only the first is a signal about conditions.

By team. This is where the variation is and where the action is. One team at three times the comparable rate is the finding.

By manager, which overlaps with team and is not identical.

By tenure. Departures in the first six months indicate a hiring or onboarding problem; departures at two to three years indicate a progression or conditions problem.

Regretted against unregretted, decided in advance rather than retrospectively.

By role and level. Losing experienced people is different from losing entry-level people, in both cost and signal.

An average across all of these is a number that cannot be acted on.

What a high rate in one team means

Before concluding anything, check the alternatives: a hot external market for that skill, a recent reorganisation, a deliberate performance exit programme, or a small team where two departures look like a crisis.

Then look at conditions: workload, vacancy duration, manager tenure, survey results for that team, hours data.

Then talk to the people still there, which is more informative than talking to the ones who left.

Why exit interviews mislead

People want a reference. Candour has no upside.

The acceptable reasons are compensation and opportunity, and those get given regardless of the real reason.

Values and manager problems are almost never stated, because doing so feels like an accusation on the way out.

The interviewer is usually seen as the organisation.

Use them, expect them to under-report conditions, and do not treat "left for more money" as an explanation.

Better instruments

Stay interviews. Ask current employees what would make them leave and what keeps them. Higher candour, and actionable while the person is still there.

Post-exit contact, several months later, by someone independent. People are markedly more candid once they have settled elsewhere.

Manager-level patterns. Several departures from one manager over a period is a signal that outranks any individual explanation given.

Where the leavers go. A cluster moving to one competitor is a different problem from a scatter.

The cost calculation

For the business case, and for prioritising which team to address first.

Recruitment cost, including agency and internal time.

Onboarding and time to full productivity, which is the largest component and the most often omitted.

Cover during the vacancy, including the load on remaining colleagues, which is how one departure produces the next.

Lost knowledge, which is hard to cost and real.

Calculate it once, properly, for your own organisation, and use it thereafter. A defensible internal figure beats any published benchmark.

The compounding effect worth naming

Departures increase the workload of those remaining, which increases the likelihood of further departures.

A team that loses two people and does not replace them promptly is at meaningfully higher risk of losing a third.

Vacancy duration is therefore a leading indicator, and it is one of the few in this field that is already sitting in a system somebody owns.

The manager pattern

The analysis that produces the most uncomfortable and most actionable finding.

Departures by manager, over a rolling period, adjusted for team size.

Compared against similar teams doing similar work in a similar market.

Read alongside that team's survey results, hours data and vacancy duration.

Handle it carefully. A single period can be noise, and there are legitimate explanations — a team of junior staff in a hot market, a deliberate performance exit, a restructure.

A sustained pattern across periods, with corroborating measures, is a finding.

Address it as a support question first: what is this manager carrying, what training have they had, what authority do they lack. Frequently the answer is structural rather than personal, and treating it as personal wastes both the person and the finding.

Post-exit contact

The instrument that produces the candour exit interviews do not.

Contact people three to six months after they left.

By someone independent, ideally not the organisation's own HR function.

Optional and clearly voluntary.

Short: what was the real reason, what would have kept you, what should we know.

Aggregate the themes and report without attribution.

Response rates are lower and candour is far higher, because the person has settled elsewhere and has nothing at stake.