Safety and Workforce KPIs: Benchmark Ranges and Targets
Recordable rates, utilization, overtime share, absence, turnover and training coverage: how to measure each KPI, what moves it, and which figures have a cited source.
Use a benchmark as a reference point for spotting a gap worth investigating, not as a target. Safety and workforce KPIs move together, so read overtime next to absence, and recordable rates next to open ergonomic findings.
Every figure shows its basis. A superscript number links to a public source in the Sources list, and the sentence gives that source's context, such as its year and the population it covers. A figure marked no cited source is kept exactly as originally published because no public source stating it has been verified, so treat it as illustrative.
At a glance
Safety and workforce KPI benchmarks at a glance
KPI
Typical
Strong
Basis
Total recordable incident rate (TRIR)
2.7 (U.S. manufacturing, 2024)
1.0 to 2.0 (strong plants); below 1.0 (world-class)
Injury rates are cases per 100 full-time workers. Words in parentheses describe each figure as the original guide or the cited source framed it, and Not stated means the guide gives no figure for that column. The cited averages are U.S. national figures; compare your plant with its own industry code first.
Method: How the figures in this guide are sourced
A figure with a superscript number is stated by the cited source, listed under Sources with its publisher, location in the document and access date. The sentence gives that source's year and population, which often differ from a single plant's. A figure marked no cited source is kept exactly as originally published because no public source stating it has been verified. It is illustrative, not measured industry data. Worked examples show arithmetic only and are not benchmarks.
Total recordable incident rate
Start safety benchmarking with the total recordable incident rate (TRIR), the headline recordable rate. In 2024, U.S. manufacturing recorded 2.7 cases per 100 full-time workers, against 2.3 for all private industry1.
World-class programs run below 1.0 and strong plants 1.0 to 2.0 no cited source. Anything above 4.0 signals a program in trouble no cited source.
Read that threshold against your own industry code. In 2024, recordable rates ranged from 0.9 in semiconductor and electronic component plants to 4.6 in forging and stamping1, so a whole subsector can average above 4.0.
How to measure
Multiply recordable cases by 200,000 and divide by employee hours worked7.
The 200,000 hours stand for 100 employees working 40 hours a week for 50 weeks a year7.
Track it monthly and as a 12-month rolling figure no cited source, so single events do not distort the trend.
What moves it
Near-miss reporting discipline.
Hazard closure speed.
Ergonomic corrective actions.
Managing to these leading indicators, because a lagging rate only moves after someone is already hurt.
Compute the rate with the TRIR Calculator, but manage to the leading indicators.
DART and lost time injury rates
The days away, restricted or transferred (DART) rate is the severity companion to TRIR. The manufacturing average runs roughly 1.7 to 2.01; the 2024 BLS rate for U.S. manufacturing was 1.7 per 100 full-time workers1.
World-class runs below 0.7 no cited source. For a published yardstick, OSHA reports that the average Voluntary Protection Programs (VPP) worksite has a DART rate 52 percent below the average for its industry8. Customer scorecards and insurers watch DART closely because it correlates with claim cost.
A DART well below your TRIR means most recordables are minor. A DART close to TRIR means your recordables are serious, and severity reduction through guarding, ergonomics and high-risk task redesign should be the priority.
Lost time injury rate is tighter still; best-in-class plants keep it under 0.5 no cited source. The closest BLS measure, the rate of cases with days away from work, was 0.8 per 100 full-time workers in U.S. manufacturing in 20241.
How to measure
Count DART cases: those with days away from work, restricted duty or a job transfer.
Use the same 200,000-hour base as TRIR, so the two rates compare directly.
Compare DART with TRIR each month to read severity.
Ergonomic risk deserves a leading-indicator KPI because sprains and strains are a large share of serious cases. Musculoskeletal disorders made up 30 percent of the cases with days away from work in U.S. private industry in 20189. That BLS figure covers all private industry, not manufacturing alone.
Score high-risk tasks and target zero jobs in the high-risk band, with a documented reduction plan for anything scoring elevated. A practical target is reassessing every manual task on a 12-month cycle no cited source and closing high-risk findings within 90 days no cited source.
Plants that hold this discipline typically see soft-tissue recordables fall 20 to 40 percent over two years no cited source.
How to measure
Score every manual task with one method and keep it fixed.
Count the jobs sitting in the high-risk band.
Track the days from finding to closure for each high-risk finding.
What moves it
Lift assists.
Workstation height.
Rotation schedules, rather than reminders to lift properly.
Labor utilization is the core workforce efficiency KPI. Typical direct-labor utilization runs 75 to 85 percent, and world-class operations reach 88 to 92 percent no cited source.
Sustained readings below 70 percent point to structural waste like material starvation or excessive changeover no cited source. The highest-leverage improvements are usually not working people harder: they are reducing waiting, cutting changeover time and fixing schedule instability.
For example, a 5-point utilization gain on a 200-person plant recovers the equivalent of about 10 full operators of capacity without hiring, since 5 percent of 200 is 10.
How to measure
Divide productive hours by paid hours.
Track it weekly for each crew.
Log the cause of each block of unproductive time, such as waiting or changeover.
Overtime as a percentage of total hours is the coverage-health KPI. A healthy target is 5 percent or below2. The closest published benchmark, in the U.S. Department of Energy's operations and maintenance guide, sets maintenance overtime under 5 percent of regular maintenance hours, so it covers maintenance staff rather than a whole plant2.
Readings of 5 to 10 percent are manageable but worth watching, and the U.S. manufacturing average sits inside that band3, 4. In August 2026 (preliminary), all manufacturing employees averaged 3.1 overtime hours in a 40.5-hour week, about 7.7 percent, and production and nonsupervisory employees 4.0 of 41.7 hours, about 9.6 percent3, 4.
Sustained readings above 10 to 12 percent signal chronic understaffing that drives fatigue, error and attrition no cited source.
Track it by crew, because a plant average can hide, for example, one department running 20 percent.
How to measure
Divide overtime hours by total hours worked.
Report it by crew and by department, not only as a plant average.
Watch it alongside fatigue and injury rates.
What moves it
Right-sizing base headcount.
Controlling absence, because overtime is usually a symptom of a base-staffing or attendance gap rather than a demand spike.
Absenteeism and turnover round out the workforce KPIs. Manufacturing absenteeism averages around 3 to 4 percent5. In 2025 annual averages, BLS put the absence rate at 2.9 percent for full-time manufacturing workers and 3.4 percent for production occupations, excluding vacation and other planned leave5.
Measured in hours rather than people, manufacturing lost 1.6 percent of worktime to absence in 20255. World-class sites hold under 2.5 percent no cited source, and every point above target directly forces overtime or short lines.
Annual turnover in manufacturing frequently runs 20 to 40 percent for hourly roles6, while best-in-class plants keep it under 12 to 15 percent no cited source.
The BLS turnover data cover all manufacturing employees, not only hourly roles. Total separations averaged 3.4 percent of employment a month in 2021 and 2.4 percent in 2025, roughly 41 and 29 percent a year when multiplied by 126. Quits alone fell from 2.4 to 1.4 percent a month over the same years, roughly 29 to 17 percent a year10.
How to measure
Divide absent people or absent hours by those scheduled, and state which basis you use.
Divide separations over the year by average headcount for annual turnover.
Report voluntary quits separately from total separations.
Track both monthly and connect the trend to overtime and staffing, so the cost of losing people stays visible to operations.
What moves it
Predictable scheduling.
Competitive pay.
Onboarding quality.
First-90-day support no cited source, since a large share of exits happen early.
Training and compliance coverage are the readiness KPIs that predict the others. Target 100 percent completion on mandatory safety training such as forklift, lockout and hazmat no cited source, and treat anything below 95 percent as an audit and incident risk no cited source.
Build the regulatory intervals into the training matrix. OSHA requires a forklift operator performance evaluation at least once every three years11 and a periodic inspection of each lockout/tagout energy control procedure at least annually12.
For audits, benchmark a compliance audit score of 90 percent or higher as pass-ready, with 80 to 89 percent needing corrective action and below 80 percent at citation risk no cited source.
How to measure
Divide required certifications held by certifications needed.
Report coverage by course, such as forklift, lockout and hazmat.
Score audits against one fixed checklist so results compare over time.
What moves it
A live training matrix that flags expiring certifications 60 days out no cited source, so coverage never lapses mid-quarter.
Regulatory re-evaluation dates entered in the matrix, not tracked from memory.
Benchmarks only drive improvement when reviewed on a cadence with owners attached. Compare against both the industry range and your own trailing 12 months no cited source, because internal trend often matters more than the external average.
Put TRIR, DART, utilization, overtime percentage, absence, turnover and training coverage on one monthly scorecard.
Give each KPI a target, an actual and a named owner.
Benchmark injury rates against your own industry code, not only the all-manufacturing average.
When a KPI drifts, trace it to a lever: rising overtime to staffing, falling utilization to changeover or material flow, climbing recordables to open ergonomic findings.
The scorecard turns isolated numbers into a management system that moves the targets year over year.
All About VPP (Voluntary Protection Programs). U.S. Occupational Safety and Health Administration. Section 'How Has VPP Improved Worker Safety & Health?' and footnote (1). Accessed September 30, 2026.