Attrition rate, and why the headline number tells you nothing
Exits divided by average headcount — a figure that is only useful once it is split by tenure, department and whether leaving was the employee's idea.
What it is
Exits over a period divided by average headcount, times 100. Usually reported annually.
The split that matters
Voluntary versus involuntary, and by tenure band. First-year exits are a different problem from fifth-year exits.
The trap
Comparing your single number to an industry average. The comparison is meaningless without the same denominator and the same definition of an exit.
Attrition rate is the percentage of employees who leave over a period: exits divided by average headcount, multiplied by 100. It is usually reported annually, and is most useful split by department, tenure band and voluntary versus involuntary.
How it works in practice
Eighteen exits against an average headcount of 120 is 15% annual attrition.
What to watch for
A single number hides everything useful. Attrition concentrated in first-year employees is an onboarding problem, not a pay problem.
How Klok handles it
Klok reports headcount and attrition from the actual joining and exit dates in the employee master.
Attrition rate FAQs
Should retirements and terminations count?
Count them, but separately. A rate that mixes resignations with retirements and dismissals measures three unrelated things and guides no decision.
What is a "good" attrition rate?
There is no useful universal figure — it varies by sector, location and role far too much. The number worth watching is your own, split by tenure, moving over time.
Why is first-year attrition tracked on its own?
Because it usually points at hiring or onboarding rather than at pay or management. Fixing it needs a different intervention entirely, so mixing it into the headline hides the signal.
How should attrition be calculated?
Consistently, over a stated period against a stated headcount base. Most disagreements about attrition are actually disagreements about the denominator.
Should we separate voluntary and involuntary exits?
Yes — combining them produces a number that cannot guide any decision. A rise driven by restructuring means something entirely different from a rise in resignations.
Which exits should be excluded from the calculation?
Be explicit either way and stay consistent. Excluding probation exits or fixed-term completions is defensible; changing the exclusions between quarters so the trend looks better is not.
Stop calculating this by hand
Klok computes it from your own attendance and salary data, every month, with the working shown.