How Do You Calculate HR Metrics Correctly, and What Do the Results Mean?

Use nine formulas: turnover, retention, first-year attrition, absenteeism, time to hire, offer acceptance, cost per hire, internal promotion and HR-to-employee ratio. Calculate each over a fixed period with a stated denominator, show the raw count beside the percentage, and read metrics in pairs so one number cannot mislead you.

I am Mac Malbrough. I hold a J.D. and an MS-HRM, I am not an attorney, and this is general information, not legal advice. This guide is for owners and HR people at 10 to 150 person organizations who can look up a formula in seconds but are less sure what to do with the answer. It pairs with my guide on which seven metrics a small employer should track, which covers the dashboard. This one covers the math and the traps.

What are the nine formulas?

1. Turnover rate: separations in the period divided by average headcount, times 100. Average headcount is start plus end, divided by two.
2. Retention rate: employees at the end of the period minus new hires in the period, divided by employees at the start, times 100.
3. First-year attrition: employees who left within 12 months of hire divided by total hires in the cohort, times 100.
4. Absenteeism rate: unplanned absence days divided by scheduled workdays, times 100.
5. Time to hire: days from the date a candidate applies, or is first contacted, to the date they accept the offer, averaged across hires.
6. Offer acceptance rate: offers accepted divided by offers made, times 100.
7. Cost per hire: internal plus external recruiting costs divided by the number of hires. The result is dollars, not a percentage.
8. Internal promotion rate: roles filled by an internal promotion divided by all roles filled, times 100.
9. HR-to-employee ratio: full-time equivalent HR staff divided by total employees, usually shown per 100 employees.

Not every metric is a percentage. Cost per hire is dollars, time to hire is days and the HR ratio is a count per 100. Label the unit on every number you share, or someone will read 38 as 38 percent.

How does a worked example look?

Here is an illustration with made-up numbers. A company starts the year with 38 employees and ends with 42. It makes 10 hires and has 6 separations. Average headcount is 40. Turnover is 6 divided by 40, which is 15 percent. Retention is 42 minus 10, which is 32, divided by 38, which is 84.2 percent. If 3 of the 10 hires leave within a year, first-year attrition is 30 percent. The company’s overall turnover looks moderate, while the first-year number says new hires are the problem. That is what pairing metrics shows.

Which denominator mistakes distort the results?

Three come up most often. First, using ending headcount instead of average headcount, which makes turnover look lower during growth and higher during a shrink. Second, mixing part-time and full-time people without converting to full-time equivalents, which skews absenteeism and the HR ratio. Third, changing the period halfway. A rolling 12 months smooths seasonality; a single month does not. Pick one method, write it on the dashboard and keep it for a year before you compare.

What do you do when your numbers are small?

At 25 employees, one resignation is a 4 percent change in turnover, and one declined offer can swing acceptance from 100 to 67 percent. Percentages on small counts exaggerate. Two fixes help. Always show the raw count next to the percentage, as in “3 of 10 offers accepted.” And use a rolling 12 months or a 24 month view for hiring metrics, so you are looking at enough events to see a pattern. Do not build a conclusion on three data points.

How should you read absenteeism in California?

Carefully, because much of what looks like absence is protected. Paid sick leave under the Healthy Workplaces, Healthy Families Act, family and medical leave, pregnancy disability leave, bereavement leave, jury duty and other protected leave are not misconduct. Labor Code 246.5 bars retaliation for using sick leave, and counting a protected absence as an attendance occurrence is generally not allowed. So calculate absenteeism only on unprotected, unplanned absences, and never use a combined absence number to discipline an individual.

Use absence data for patterns, not for blame. A team with a rising absence count may have a scheduling, workload or management problem. That is a signal to investigate the conditions, not the people.

What does a low offer acceptance rate usually mean?

Usually one of four things: pay below the market for the role, a slow process that let the candidate accept somewhere else, a mismatch between what the posting promised and what the interview revealed, or a weak close. Check them in that order. California requires employers with 15 or more employees to include a pay scale in job postings (Labor Code 432.3), so a mismatch between the posted range and the actual offer is easy for a candidate to notice. Pair acceptance with time to hire: a long time to hire plus a low acceptance rate almost always means you are losing candidates to speed.

How do you use cost per hire without gaming it?

Count everything: job board fees, recruiter fees, referral bonuses, the time managers spend interviewing and the cost of the onboarding setup. Then compare by role, because an administrative hire and a specialist hire do not belong in one average. A low cost per hire is not automatically good. If you hired cheaply and the person left in four months, the real cost was higher. This is why first-year attrition and cost per hire belong on the same page.

Does the internal promotion rate prove you develop people?

It suggests it, but check the quality. A high rate with weak outcomes can mean you promote people because it is easy, not because they are ready. A low rate can be healthy in a small company with few steps in the ladder. Read it with the first-year performance of promoted employees and with exit interview themes about growth. Then compare against your own history, not against a general benchmark that does not match your size.

What is a reasonable HR-to-employee ratio for a small employer?

It depends on the mix of work, the number of locations and how much is outsourced. Published benchmarks vary widely by industry, size and what counts as HR, so I treat any single figure as a conversation starter, not a target. In employers under 50, HR is often a fractional role, so the ratio will read low even when the coverage is adequate. Use the ratio to ask whether HR work is getting done, not to decide whether you have the “right” number of people.

How do you explain the results to an owner?

Lead with the decision, not the formula. Instead of “turnover is 15 percent,” say “six people left this year, three of them within their first year, and we need to fix how we onboard.” Show the number, the raw count, the direction compared with last period and the one action you recommend. Owners rarely want the math; they want to know whether to worry and what it will cost. Keep the formulas on a second tab for anyone who asks.

What are the signs that a metric is misleading you?

Watch for four. The number moves sharply with no change in the business, which usually means a counting change. The metric is good while people are unhappy, which means it is measuring the wrong thing. Everyone agrees on the number and no one acts on it, which means there is no threshold. Or the number improves because the definition changed. If any of these show up, go back to the denominator and the period before you trust the trend.

How often should you calculate these?

Quarterly for most, and monthly for hiring metrics when you are recruiting. Put the formulas in a shared spreadsheet, keep the raw counts on a second tab and write one sentence of interpretation for each number. Then decide before the meeting what number would make you act. A metric with no action threshold is a decoration.

How does the UHR pay equity workbook connect to these metrics?

The problem with HR numbers is that a turnover rate or a time-to-hire figure tells you something is off but not where. Pay is one of the places to look, and it is the one most likely to carry legal risk. The Pay equity audit workbook ($199) gives you the spreadsheet structure, comparator-group logic and remediation-plan template, so a pay question raised by your metrics becomes a documented review instead of a guess. For the metrics themselves, the formulas above are all you need to start.

What should you do this week?

Pick three metrics, turnover, first-year attrition and offer acceptance, and calculate them for the last 12 months with raw counts beside the percentages. Write one sentence on what each tells you and one action you would take if it moved the wrong way. If you want a ready-made structure, the pay equity workbook shows how I lay out data tabs, and a compliance check-up can review the records behind the numbers.

Related reading: how to show the return on employee engagement.

For a related angle, see people analytics terms in plain English and building a reusable HR reporting workbook.

Sources

• California Labor Code Section 246.5, read on the California Legislative Information site
• California Labor Code Section 432.3, read on the California Legislative Information site

McKinley holds a J.D. but is not a licensed attorney. Articles here are general information, not legal advice. For your specific situation, talk with an employment attorney.

Data current as of November 2026. Sources are linked where each figure appears.