Retention Rate

How to calculate retention rate correctly

Freeze a list of employees on the first day of the period. At the end, count how many people on that list are still employed. Divide by the starting count. That is the whole method, and its strength is that it ignores hiring entirely.

Two decisions change the answer, so settle them before reporting:

• Internal moves. Someone who transferred from the warehouse to the dispatch office is retained. Match on employee ID, not department roster.
• Leaves of absence. People on approved leave are still employed. Count them as retained.

For a quarterly or monthly rate, use the same method over the shorter window. Retention over shorter windows is naturally higher, so always label the period.

Retention rate is not 100 minus turnover

This is the most common error on HR dashboards. Employee turnover divides all separations, including new hires who quit in week two, by average headcount. Retention looks only at the starting group.

Using the hospital example: total separations were 52 + 14 = 66, and average headcount was about 178 (180 at the start, 175 at the end), giving turnover of about 37%. Retention was 71%. Those do not add to 100, and they should not. A site with heavy early attrition can show a decent retention rate for its tenured core and an ugly turnover rate at the same time.

That gap is useful. It tells you whether your problem sits with long-tenured staff or with the people you just hired.

New hire and cohort retention

For high-volume employers, the most useful version of this metric is cohort retention: take everyone who started in a given month and track what share is still employed at 30, 90 and 180 days.

A contact center that hires 40 agents into a February class might see 36 still employed at day 30 (90%), 31 at day 90 (78%) and 26 at day 180 (65%). Plot each monthly class on the same chart and changes jump out: if the May class holds 85% at day 90 and the June class holds 66%, something changed in sourcing, the recruiter team, the trainer or the schedule offered.

Cohort retention is the mirror image of new hire turnover and is the fairest way to judge whether recruiting and onboarding changes worked.

Typical retention ranges

Annual retention for salaried professional staff is commonly 85% to 92%. Hourly retail, food service and warehouse roles often sit between 40% and 70%, depending on the local labor market and pay. For new hourly hires, 90-day retention of 70% to 80% is typical, and strong programs push it above 85%.

Treat these as rough guides. Retention among your top performers matters more than the average, so cut the figure by performance rating as well as by site and role.

What moves retention

Retention is mostly decided in three places.

The hiring decision. People who knew what the job involved before accepting stay longer. Realistic previews, clear shift information and honest pay conversations cost nothing and raise 90-day retention.

The first month. A structured onboarding plan, a named buddy and a supervisor check-in at day 7 and day 30 catch problems while they are still fixable.

The direct manager. Retention by supervisor is often the most revealing cut of all. Two team leads running the same work with the same pay can have retention rates 20 points apart.

Pay matters too, but it rarely works alone. A modest raise rarely keeps someone whose schedule changes every week. For the wider program view, see employee retention and run regular conversations with our stay interview template.

Common mistakes

• Including new hires in the numerator. It inflates the rate and hides early attrition.
• Counting transfers and promotions as exits. Internal movement is retention.
• Reporting one company-wide rate. Averages hide the two sites or three supervisors driving most of the losses.
• Ignoring who stayed. A 90% retention rate where the top performers left is worse than an 80% rate where they stayed.
• Measuring once a year. By the time an annual figure shows a drop, the people are gone. Review cohort data monthly.