Regretted Attrition
Turnover the company actively wanted to prevent — a strong performer or valued employee leaving — as distinct from non-regretted attrition, which the organization is fine with or even welcomes.
Regretted attrition is the subset of employee turnover an organization specifically didn't want to happen: a high performer, a hard-to-replace specialist, or someone central to team functioning leaving voluntarily. Non-regretted attrition is the mirror category — departures of underperformers, poor culture fits, or people in roles being eliminated anyway — where the company's overall talent position is neutral or even improved by the exit. The distinction matters because a company's raw turnover percentage tells you almost nothing on its own; a 15% turnover rate composed mostly of regretted departures is a serious problem, while the same 15% composed mostly of non-regretted departures might reflect healthy performance management.
HR and people analytics teams track regretted attrition as a leading indicator of organizational health, often broken down further by team, tenure band, and performance rating, to isolate where the company is actually losing people it wanted to keep. A spike in regretted attrition concentrated in one team, or among a specific tenure band (commonly 18–30 months, past onboarding but before deep investment pays off), is a much more actionable signal than an aggregate turnover number, because it points at a specific root cause — a manager, a compensation gap, a stalled growth path — rather than a vague company-wide trend.
The classification itself requires a judgment call, usually made by the departing employee's manager shortly after they announce their departure, and it's an imperfect measure — managers can be reluctant to classify a departure as regretted if it reflects poorly on their own leadership, or may mischaracterize a low performer's exit as regretted to avoid the appearance of a performance management failure. Organizations that use the metric well typically calibrate these classifications across managers, similar to performance rating calibration, rather than taking each manager's self-reported classification at face value.
Regretted vs. Non-Regretted Attrition
- Regretted: a valued employee — strong performer, hard-to-replace specialist, key team contributor — leaves voluntarily, and the company would have preferred to keep them.
- Non-regretted: an underperformer, a poor culture fit, or someone in a role being phased out leaves, and the company's overall position is neutral or improved.
- Aggregate turnover rate alone doesn't distinguish between these — two companies with identical headline turnover numbers can be in very different positions depending on the regretted/non-regretted split.
Why the Split Matters More Than the Headline Number
A rising regretted attrition rate concentrated in a specific team, tenure band, or performance tier points directly at a fixable root cause — a struggling manager, a compensation gap relative to market, or a stalled growth path — in a way that an aggregate turnover percentage never can. Organizations that only track the headline number risk missing a serious, specific retention problem simply because it's diluted by unrelated, unproblematic departures elsewhere in the business.
Example
A 40-person engineering org has 12% annual turnover, which sounds moderate. A deeper look shows that 8 of the 12 percentage points are regretted — including three of the team's five most senior engineers, all citing stalled promotion timelines in their exit interviews. The other 4 points are non-regretted departures the org was fine to see go. Leadership treats the 8% regretted figure, not the 12% headline number, as the real signal — and it directly triggers a review of the engineering promotion process.