Retention is not one hundred minus turnover. The two use different denominators and answer different questions, and this calculates the cohort version properly. Every number on this page is calculated in your browser and never leaves your device.
A cohort is built to be followed forward, which makes retention the one metric that is almost useless measured once. The HRAnalyst™ Workbook is a printable tracker for every calculator on this site, with the formula, the recommended cadence and space for four quarters. Join the list and we will send it, along with new calculators and workforce analytics as they publish.
Formula used
Retention Rate = (Still Employed from Cohort ÷ Cohort Starting Count) × 100
Employees Lost from Cohort = Cohort Starting Count minus Still Employed from Cohort
Turnover Rate = (Separations ÷ Average Headcount) × 100
The third line is the turnover formula, shown here so the difference is on the page rather than in your head. Retention divides by a cohort, a fixed list of named people counted once at the start. Turnover divides by an average headcount, a moving population that includes everyone who joined partway through. Two different denominators, so the two rates do not add up to one hundred except by coincidence.
Retention and turnover are not opposites. They divide by different things, so subtracting one from one hundred does not produce the other. Here is what each one is actually counting.
The denominator is the cohort you counted on day one, and it never changes. It is a list of named people. Anyone hired after the window opened is outside that list and cannot affect the rate, which is exactly what makes retention a clean read on whether the people you already had chose to stay.
Turnover divides departures by an average headcount that includes people who arrived after the period began. The denominator grows as you hire, which dilutes the rate even when nothing about the departures changed. That is not a flaw in turnover. It is simply a different question.
An organization hiring quickly can post good turnover and poor retention in the same quarter, and both numbers are correct. When someone in the room insists the two figures contradict each other, the answer is that they are answering different questions. Say which denominator each one used and the disagreement usually ends there.
The subtraction is quick and it is wrong. The two rates use different denominators, so the result is a number with no definition behind it. If you report it and someone recalculates retention properly, the figures will not match and you will have to explain why.
The cohort is by definition who was there at the start. Adding later hires inflates the denominator with people who had less time to leave, and the rate improves for a reason that has nothing to do with retention. Pull the list as of the first day of the window and freeze it.
A very short window flatters every cohort, because leaving takes time. If you report a high retention figure on a narrow window, state the window in the same sentence as the number so nobody reads it as an annual result.
The calculators and content on HRAnalyst are provided for informational purposes only and do not constitute legal, tax, or accounting advice. Compensation and compliance obligations vary by jurisdiction. Consult qualified counsel before acting on any result.
© 2026 HRAnalyst is a product of CompRatio LLC. All rights reserved.
Looking for salary data? Visit What It Pays™.