Regretted Attrition

A skilled technician resigns from a role with few trained replacements, and management may classify that exit as regretted and look for preventable reasons. The point is to learn about workforce risk, not to dismiss other employees' departures as unimportant.

SHRM discusses assessing the quality of attrition and Visier examines reasons employees resign, but neither supplies a universal definition of which individuals a business must label "regretted", so a fair, documented method matters. Define the category before reviewing a year's results, because otherwise managers can reclassify exits to fit a target, and review the criteria again afterward since a manager may underclassify regretted exits to look good.

State whether it includes only voluntary resignations or also other separations; the definition here focuses on voluntary departures. Use criteria relevant to work and continuity, such as needed expertise or documented performance, not a manager's vague personal preference, and check for bias across teams and demographic groups where lawful, since subjective labels can reproduce unfair views.

For an annual rate, divide regretted voluntary leavers during the year by average eligible headcount and multiply by 100, so if 12 such employees leave from average headcount of 200, the illustrative rate is 6%. Record the count as well as the rate, because in a small team a single resignation can make the percentage swing sharply.

Set a useful period too, since month-to-month percentages can be noisy for small organisations. Do not assume every departure could have been prevented, since a person may leave for reasons outside the employer's control.

Ask employees about conditions while they remain, through appropriate stay conversations or surveys, because exit interviews can provide clues but may not reveal the full reason when some departing workers do not feel safe being candid. Look at manager support, growth, pay, workload and flexibility alongside available evidence, and avoid claiming a cause from one anecdote.

Track critical-role coverage and succession, since even a low rate can be serious if a unique specialist leaves. Compare by role, tenure and team while protecting privacy, because small cells can identify individuals, and review the time and cost to replace lost capabilities since a headcount rate alone does not show disruption.

Separate planned turnover from unexpected departures in succession reporting, and track voluntary attrition overall as well, because a falling regretted rate could hide a large rise in other exits. A departing employee who joins a better opportunity may still have had a good experience, so the company's regret is not theirs to carry and nobody should be pressured to stay after a decision.

Monitor internal mobility and career paths, because talented people sometimes leave when they cannot see another role inside, and compare pay policies with external markets carefully since a one-off counteroffer may not fix broader fairness or growth concerns. Do not equate high performance ratings with every valuable contribution, share aggregate insights with leaders who can improve conditions rather than a public list of "valuable" and "not valuable" leavers, and mark any definition change after a reorganisation before comparing trend lines.