How to measure safety program effectiveness for small business

Last updated 2026-07-11

Why measuring your safety program matters more than writing it

A lot of small business owners treat safety programs like a filing cabinet project. Write the program, put it in a binder, move on. That binder does nothing for you unless you can answer one question: is it actually working?

Most small employers can't answer it. They find out their program failed the hard way. Someone gets hurt, a workers' comp premium jumps, or an OSHA inspector shows up asking questions the binder never anticipated.

Measurement flips that. Track the right numbers and you see problems early enough to fix them. A rising near-miss count. A training completion rate that keeps slipping. An audit finding that never gets closed out. These are warning signals that show up weeks or months before an injury does.

The Bureau of Labor Statistics counted 2.6 million nonfatal workplace injuries and illnesses in private industry in 2023 ^[1]. Small businesses, usually defined as fewer than 500 employees, carry a heavy share of that number because they rarely have someone whose job is watching the data. Measurement is how you cover for not having a full-time safety manager.

None of this needs expensive software or a consultant. You need a spreadsheet, a schedule you actually keep, and maybe two hours a month.

What is TRIR and how do you calculate it for a small business?

TRIR stands for Total Recordable Incident Rate. It's the most widely used lagging indicator in occupational safety, and it's the number OSHA, insurance carriers, and general contractors use to size up your safety record against the rest of your industry.

The formula is short:

TRIR = (Number of OSHA recordable injuries and illnesses x 200,000) / Total hours worked

The 200,000 figure represents 100 employees working 40 hours a week for 50 weeks. It's the standard base so companies of different sizes compare on even footing ^[2].

Say you run a 12-person landscaping crew. Last year they worked a combined 22,000 hours and had 2 recordable injuries. Your TRIR is (2 x 200,000) / 22,000 = 18.2. That's high. The BLS average for landscaping services (NAICS 5617) has run roughly 3 to 4 in recent years ^[1]. That gap tells you something is wrong, and it tells you so with a number instead of a hunch.

Here's the catch for the smallest employers. With 10 or fewer people, one injury can wreck your rate even if your program is solid. That's just math on a tiny sample. Track the rate anyway, but read it over two or three years instead of month to month. One injury in a year with 8,000 hours gives you a TRIR of 25. Add a person and 2,000 more hours the next year with zero incidents and your rolling two-year rate falls to 10. The trend line is the real signal.

To find your benchmark, pull the BLS injury and illness data tables and find your NAICS code ^[1]. Compare your TRIR to the total recordable cases rate for your sector. Above average means work to do. At half the industry average, your program is working.

What other injury rate metrics should small businesses track?

TRIR counts every recordable event but says nothing about how bad your injuries are. For severity you need two more rates.

DART rate (Days Away, Restricted, or Transferred) counts only cases serious enough to send a worker home, put them on restricted duty, or move them to another job. Same formula as TRIR, smaller numerator. A high DART next to a modest TRIR means your injuries are severe when they happen, which points to different fixes than a high TRIR made mostly of borderline recordables.

DAFW rate (Days Away From Work) is a narrower cut still, counting only cases with actual lost workdays. The BLS publishes benchmark data for both DART and DAFW rates by industry ^[1], so all three metrics have a national yardstick.

Metric	What it measures	BLS benchmark available?
TRIR	All recordable injuries per 100 FTE	Yes, by NAICS
DART	Serious cases per 100 FTE	Yes, by NAICS
DAFW	Lost-time cases per 100 FTE	Yes, by NAICS
Near-miss rate	Reported close calls per 100 FTE	No standard benchmark
First aid rate	Minor incidents per 100 FTE	No standard benchmark

First aid cases, the ones you treat on-site and never enter on the OSHA 300 log, are worth tracking informally. A high first-aid-to-recordable ratio can mean your controls are catching things early. It can also mean people are shrugging off injuries they should report. Read it in context.

New to OSHA recordkeeping? The incident report article walks through what counts as recordable under 29 CFR 1904 and what doesn't.

What are leading indicators and why do they matter more than injury counts?

Lagging indicators like TRIR look backward. They tell you what already went wrong. Leading indicators look forward. They measure the activities and conditions that predict whether an injury will happen.

This is the part most small businesses skip. It's also where the real measurement payoff lives.

Here are the leading indicators worth tracking:

• Near-miss reports per month. A near-miss is an event that could have caused injury but didn't. The old Heinrich triangle argued that many near-misses and unsafe acts precede every serious injury. The exact ratios have been picked apart over the years, but the direction holds: near-misses flag real hazards. If your monthly count is zero, that's almost never good news. It means people aren't reporting, not that your shop is flawless.

• Safety observation rate. How many times a week does a supervisor actually walk the floor and write down what they see? Zero means nobody is looking.

• Training completion rate. What share of your required training is current? If you run powered industrial trucks, 29 CFR 1910.178(l) requires forklift certification training before operation and re-evaluation at least every three years ^[3]. If 30% of your operators are overdue, that's a leading indicator that's already a live compliance gap.

• Corrective action close-out rate. When an audit or inspection finds a hazard, how long until it's fixed? Average days to close is a brutally honest number. It tells you whether your program has teeth or just paperwork.

• Safety meeting attendance rate. If nobody shows, the program lives on paper only.

None of these need a software system. A spreadsheet updated weekly covers most businesses under 50 employees. The work is looking at the numbers on a schedule and asking what they mean.

How do you run a safety audit and score it consistently?

A safety audit is a structured inspection of your workplace against a defined checklist. Structured is the operative word. A walk-around where someone notes what looks bad is not an audit. An audit produces a score you can track over time.

Start by building a checklist across your main hazard categories: housekeeping, emergency exits, PPE availability and condition, machine guarding, chemical storage, lockout/tagout if you have stored-energy hazards, and electrical safety. For each item, mark it compliant, non-compliant, or not applicable. Score it as the percentage of compliant items out of applicable items.

40 applicable items with 32 compliant gives you an audit score of 80%. Run the same audit quarterly. A climb from 80% to 85% to 91% over three quarters means the program is improving. Flat or falling means you're not closing last quarter's findings.

For hazard communication, 29 CFR 1910.1200 requires that your written program, your SDS collection, and your employee training all exist and be verifiable ^[4]. That's three separate audit checkpoints in one standard.

OSHA's small business resources include inspection checklists organized by industry ^[5]. Use those as your template instead of starting cold. The OSHA On-Site Consultation Program, which is separate from enforcement and free to small employers, will also walk your facility and hand you a hazard list ^[5]. Treat that report as your audit baseline.

Frequency tracks your hazard level. A low-hazard office might audit once a year. A fabrication shop or job site should run monthly. Most small manufacturers land around quarterly formal audits with informal supervisory walk-throughs each week.

How do you track near-miss reporting and why is the number often zero?

Near-miss reporting is the metric most predictive of future injuries, and the one most reliably broken at small businesses.

Zero reports rarely means a perfectly safe workplace. It usually means workers don't report because they expect paperwork, blame, or embarrassment. If the first person who reports a near-miss gets a lecture or a disciplinary form, nobody reports again.

You need a no-blame system. That can be a box on the wall with paper forms, a shared email address, or a text to a supervisor. The form should ask four things: what happened, where, what the hazard was, and what would fix it. No name required unless the person wants to add one.

Track reports per month. Fifteen employees and zero near-miss reports in three months means your reporting system is broken, full stop. A rough target for a small manufacturing or construction crew is about one near-miss report per 10 workers per month, though nobody has clean benchmark data on this. What you want is a nonzero number that climbs as your reporting culture warms up.

When a near-miss lands, document it, find the root cause, fix the hazard, and tell the crew what you did. That feedback loop is what gets you the next report.

What does workers' comp cost data tell you about safety performance?

Your workers' compensation experience modifier (the e-mod, or experience modification rate) is a direct financial read on your injury history against other businesses in your industry and state. An e-mod of 1.0 is average. Below 1.0 means your injury costs run lower than average and you pay less in premium. Above 1.0 means the reverse.

Your state's workers' comp rating bureau calculates the e-mod from three years of your claims history, usually leaving out the most recent policy year. It isn't an OSHA metric, but it's one of the most honest lagging indicators you have, because it's built on real claim dollars rather than on whether an injury cleared the recordable bar.

An e-mod of 0.80 means you pay 20% less than the industry average. An e-mod of 1.30 means 30% more. The National Council on Compensation Insurance (NCCI) runs the e-mod system in most states ^[6]. Ask your broker for your current e-mod and its three-year trend. That trend line reads out your safety program's financial effect better than almost anything else.

The indirect costs of an injury dwarf the direct ones. OSHA estimates that for every dollar of direct workers' comp cost, employers pay $1.15 to $4.50 in indirect costs: lost productivity, overtime, training a replacement, administrative time ^[7]. Tracking total incident cost, more than premium impact, gives you the full picture.

How do you measure whether safety training is actually working?

Training completion tells you someone sat through a session. It says nothing about whether behavior changed.

Measure at two levels. First, knowledge retention: can workers correctly answer questions about the hazard and the procedure? A short quiz right after training, then again 30 to 60 days later, shows how much stuck. Second, behavior change: are workers actually using the safe procedure on the floor? Supervisory observation is the only way to see this.

The regulatory floor changes by standard. OSHA training requirements under standards like 29 CFR 1910.178 (powered industrial trucks), 29 CFR 1910.147 (lockout/tagout), and 29 CFR 1910.134 (respiratory protection) each spell out who gets trained, on what, and how often ^[3]. Meeting those minimums is the starting line, not the finish.

For supervisors and managers, an OSHA 30 course gives a deeper grounding in spotting hazards and understanding OSHA obligations ^[8]. That investment tends to show up in better audit scores and higher near-miss reporting over the following year, though I can't point you to a controlled study that pins down the effect size.

Track two numbers. Completion rate (the share of required training that's current) and pass rate on post-training assessments. Anything below 90% completion on required training is a compliance problem. Pass rates below 80% mean the content or the delivery needs work.

How often should you review your safety metrics and who should see them?

Metrics nobody looks at are dead weight. Set a review cadence before you collect a single number.

Monthly: review near-miss reports, safety observation counts, training completion, and any open corrective actions. Thirty minutes if your data is organized.

Quarterly: run the formal safety audit and update your TRIR and DART calculations. Compare against the same quarter last year. Review workers' comp claim activity with your broker.

Annually: pull BLS benchmark data for your NAICS code and compare your rates. Check whether your e-mod moved. Read your written safety program for accuracy and update it if your operations, equipment, or hazards changed during the year.

Who sees the numbers matters as much as the numbers. Post your TRIR and near-miss counts where workers can see them. Not as a threat, but as proof you're paying attention and that their reports drive real change. Workers who watch their near-miss reports turn into fixed hazards report more near-misses. That loop is how safety culture actually builds.

Still working on getting the written program itself in order? SafetyFolio's safety program generator builds a complete, OSHA-aligned program in about 15 minutes, which gives you the documented foundation these metrics are supposed to grade.

What's a simple scorecard a small business can use starting today?

You don't need a dashboard platform. One spreadsheet, seven columns, updated monthly, covers most small businesses.

Metric	How to calculate	Target
TRIR	(Recordables x 200,000) / Hours worked	Below your BLS industry average
DART rate	(DART cases x 200,000) / Hours worked	Below your BLS industry average
Near-miss reports	Count per month	Increasing, nonzero
Training completion	% of required training current	100%
Audit score	% of items compliant	Improving quarter over quarter
Corrective action close rate	% closed within 30 days	Above 90%
Safety meeting attendance	% of employees attending	Above 90%

Score each metric green, yellow, or red against its target. Review monthly. Three reds at once means a systemic problem, not a random bad stretch.

The BLS publishes its industry injury rate tables once a year, usually in the fall for the prior year ^[1]. Set a calendar reminder to refresh your benchmarks. Aim first to reach and hold below the median for your NAICS code, then push toward the lower quartile.

For a shop with fewer than 10 people, run this review in a short monthly team meeting. Reading the numbers out loud and asking "what do we think drove this?" produces more useful insight than any consultant report.

What are common mistakes small businesses make when measuring safety?

The biggest one is measuring only after something bad happens. If your first safety metric is a TRIR you calculate the week after someone breaks a wrist, that injury is already gone.

Second: measuring only lagging indicators. TRIR and DART matter, but if those are the only numbers you watch, you have no warning system. Pair them with two or three leading indicators.

Third: underreporting. Leaning on workers not to report injuries because it makes the numbers look bad is illegal under 29 CFR 1904.36, OSHA's anti-retaliation provision ^[9], and it's self-defeating. Suppressed data wrecks your ability to find real hazards. An unrealistically low TRIR at a facility with obvious physical hazards is a red flag to inspectors, not a trophy.

Fourth: measuring things you never act on. If audits keep flagging the same unguarded belt drive and nothing changes, your audit is theater. The corrective action close rate in the scorecard above exists to catch exactly this.

Fifth: ignoring near-misses. OSHA's Voluntary Protection Programs treat near-miss investigation as one of the core practices that separates strong safety programs from average ones ^[10]. You don't have to be a VPP site to steal the practice.

If your state runs its own OSHA-approved plan, some requirements go past the federal minimum ^[11]. California's Cal/OSHA, for one, requires an Injury and Illness Prevention Program (IIPP) with specific recordkeeping and review elements for nearly all employers. Check whether your state adds measurement or documentation requirements on top of the federal baseline.

How does OSHA use injury rate data during inspections?

OSHA uses injury rate data in a few direct ways that reach small businesses.

First, OSHA's Site-Specific Targeting program uses injury and illness data submitted by establishments to pick high-rate workplaces for programmed inspections ^[12]. Under the Electronic Recordkeeping Rule (29 CFR 1904.41), establishments with 100 or more employees in certain high-hazard industries must submit their 300A data electronically each year, and establishments with 20 to 99 employees in listed industries submit as well. Smaller establishments face lower reporting duties, but OSHA can still use available data to set priorities.

Second, during an inspection, compliance officers start with your OSHA 300 log, 300A summary, and 301 incident report forms. If the log shows a pattern, say multiple strains from the same task or a cluster of injuries in one department, the inspector digs into why. A well-kept log with documented corrective actions tells a far better story than a blank log that looks like injuries simply aren't being recorded.

Third, your injury data can affect proposed penalties. In OSHA's penalty calculation, a demonstrated good-faith effort to find and fix hazards is one of the factors that can lower a proposed penalty ^[13]. Documented safety metrics, closed corrective actions, and training records are your evidence of that good faith.

Want the broader picture of what OSHA can and can't do on an inspection? The osha overview covers the agency's enforcement authority in plain language.

Frequently asked questions

What is a good TRIR for a small business?

Good depends entirely on your industry. Compare your TRIR to the BLS average for your NAICS code. Below the average is a pass; at half the average or less is genuinely strong. For all private industry combined, BLS reported a total recordable case rate of 2.4 per 100 full-time workers in 2023. Construction, manufacturing, and agriculture run higher. Check the annual BLS injury and illness report for your specific sector.

Do small businesses with fewer than 10 employees have to keep OSHA injury records?

Employers with 10 or fewer employees are partially exempt from OSHA's routine injury recordkeeping under 29 CFR 1904.1. They don't have to keep a 300 log unless OSHA or BLS specifically requests it. They still must report any work-related fatality within 8 hours and any inpatient hospitalization, amputation, or loss of an eye within 24 hours. The recordkeeping exemption doesn't mean injuries stop mattering for measurement.

What is a near-miss report and how is it different from an incident report?

A near-miss is an unplanned event that could have caused injury or property damage but didn't. An incident report documents an event where injury or damage actually happened. Near-misses don't go on your OSHA 300 log since no injury occurred, but tracking them internally is one of the most useful leading indicators you have. They reveal real hazards before anyone gets hurt, which makes them more actionable than incident data.

How do I find the BLS injury rate benchmark for my industry?

Go to the BLS website and find the annual Survey of Occupational Injuries and Illnesses data tables. Look for the table covering your industry sector. You'll need your NAICS code, which appears on your IRS business filings or on the Census Bureau NAICS search tool. BLS publishes total recordable case rates, DART rates, and days-away-from-work rates by industry each year for the prior calendar year.

Can OSHA penalize a small business for having a high injury rate alone?

A high injury rate by itself doesn't trigger a penalty. OSHA cites specific standards violations, not injury counts. But a high rate can trigger a programmed inspection under OSHA's targeting programs, and during that inspection violations may be found and cited. The rate is a flag that gets you looked at more closely, not a fine on its own.

What leading indicators should a small construction company track?

Strong construction leading indicators include toolbox talk attendance rate, fall protection inspections completed before each shift at height, the share of new workers who got site-specific orientation before starting, and near-miss reports per week. Falls account for the largest share of construction fatalities, so any metric tied to fall protection inspection and training completion is especially predictive of serious injury risk.

How often should a small business run a formal safety audit?

For most small businesses in moderate-to-high hazard industries (construction, manufacturing, warehousing), quarterly formal audits with informal weekly walk-throughs is a reasonable cadence. Low-hazard offices can often do one formal audit a year. Consistency is the point: run the same checklist each time so you compare apples to apples and track your score over time instead of just noting what looks bad on a given day.

What is an experience modifier and how does it connect to safety performance?

The experience modifier (e-mod) is a multiplier your workers' comp insurer applies to your premium based on three years of claims history compared to your industry average. An e-mod of 1.0 is average. Below 1.0 means you pay less; above 1.0 means you pay more. It's a direct financial signal of your safety program's effect. Ask your broker for your current e-mod and how it has trended over the past three years.

Is it illegal to pressure employees not to report injuries to keep TRIR low?

Yes. 29 CFR 1904.36 prohibits employers from retaliating against workers for reporting work-related injuries and illnesses. OSHA's anti-retaliation rule also bars policies that discourage reporting, including incentive programs that reward injury-free periods in ways that create implicit pressure not to report. Violations can bring citations and penalties separate from any underlying injury standard violation.

What free resources does OSHA offer to help small businesses measure and improve safety?

OSHA's On-Site Consultation Program provides free, confidential workplace hazard assessments for small and medium businesses. Consultants are separate from enforcement staff and cannot issue citations. The program can also help you identify leading indicators and set up basic measurement. Find your state's consultation program through OSHA.gov. The program is OSHA-funded but delivered through state agencies, so services vary a little by state.

How do safety metrics differ for businesses in states with their own OSHA plans?

The 29 states and 2 territories with OSHA-approved plans must meet or exceed federal OSHA requirements. Some go further. California's Cal/OSHA, for example, requires all employers to have a written Injury and Illness Prevention Program that includes a system for investigating incidents and documenting corrective actions, which is a mandatory measurement element. If your state runs its own plan, check with your state OSHA agency for extra recordkeeping or program review requirements beyond the federal baseline.

Can a small business with zero injuries still have a bad safety program?

Absolutely. Zero injuries in a year can mean your program is excellent, or that you have a small crew with few exposure hours and got lucky, or that injuries are happening and not being reported. Zero injuries alongside zero near-miss reports, poor audit scores, and overdue training is not a safe operation. It's an operation that hasn't had its first serious injury yet. Leading indicators show you the real picture.

How long does it take to set up a basic safety measurement system?

Realistically two to four hours to set up, then one to two hours a month to maintain. You need a spreadsheet with the seven core metrics, your OSHA 300 log (required for most employers with 11 or more employees), a near-miss reporting form, and an audit checklist for your main hazard categories. The OSHA On-Site Consultation Program and BLS benchmark tables are free. The limiting factor is the discipline to review monthly, not setup complexity.

Sources

1. Bureau of Labor Statistics, Survey of Occupational Injuries and Illnesses: 2.6 million nonfatal workplace injuries and illnesses in private industry in 2023; industry-specific TRIR, DART, and DAFW rates by NAICS code published annually
2. OSHA, Injury Tracking Application and recordkeeping guidance: TRIR formula uses 200,000 as the base (100 employees x 40 hours x 50 weeks) for industry comparison
3. OSHA, 29 CFR 1910.178 Powered Industrial Trucks standard: Requires operator training before use and evaluation at least every three years under 29 CFR 1910.178(l)
4. OSHA, 29 CFR 1910.1200 Hazard Communication standard: Requires written hazard communication program, safety data sheets, and employee training on chemical hazards
5. OSHA, On-Site Consultation Program for small businesses: Free, confidential on-site consultation separate from enforcement; available to small and medium businesses in all states
6. National Council on Compensation Insurance (NCCI), experience rating overview: NCCI administers the experience modification rate system used in most states; e-mod below 1.0 results in premium reduction
7. OSHA, Business Case for Safety and Health: OSHA estimates indirect costs of injuries run $1.15 to $4.50 for every $1.00 of direct workers' comp cost
8. OSHA, Outreach Training Program (OSHA 10 and 30): OSHA 30-hour course provides in-depth hazard recognition and OSHA compliance training for supervisors and workers
9. OSHA, 29 CFR 1904.36 Prohibition against discrimination: 29 CFR 1904.36 prohibits retaliation against employees for reporting work-related injuries or illnesses
10. OSHA, Voluntary Protection Programs (VPP) overview: Near-miss investigation is identified as a core practice in OSHA VPP guidance for high-performing safety programs
11. OSHA, State Plans overview: 29 states and 2 territories operate OSHA-approved state plans that must meet or exceed federal OSHA standards; some impose additional requirements
12. OSHA, Enforcement and inspection programs: OSHA's Site-Specific Targeting program uses submitted injury and illness data to select high-rate establishments for programmed inspections
13. OSHA, Penalties: Good faith effort to comply with OSHA requirements is one of the factors OSHA uses to reduce proposed penalty amounts