High BBS observation rates with flat incident numbers aren't a paradox — in manufacturing, they're usually a turnover problem wearing a data mask.
Your BBS observation rate looks great on paper. Observers are hitting their weekly card quotas, the safe-to-at-risk ratio is sitting around 80/20, and your safety coordinator is showing the numbers to plant leadership every month. And yet — your TRIR hasn't moved in two years. Maybe it's actually crept up.
This is one of the most disorienting positions a safety manager in a mid-size manufacturing plant can land in. The leading indicator is green. The lagging indicator doesn't agree. Something is broken — and it's almost never what people initially assume it is.
In a stable workforce — one where your press operators, maintenance techs, and material handlers have been on the floor for two, three, five years — a BBS observation program can actually work as designed. The observer knows the job. The person being observed knows the observer. The conversation after the card is filled out has history behind it. That peer-to-peer dynamic is what drives behavior change, not the card itself.
Now run 40% turnover through that plant in eighteen months. Add a temp agency feeding you twelve new production workers every quarter. Layer in a subcontractor crew running a capital equipment installation in Bay 4 while your regular millwrights work around them.
What happens to your observation program? The cards still get filled out — because the quota didn't change and the safety coordinator still needs the numbers for the monthly report. But the observer is now watching a temp worker they've known for six days perform a task on equipment they've never operated before. The observer doesn't know what "normal" looks like for that job. They check "safe" on lifting technique and move on.
Your observation rate stays high. Your data quality collapses. And because the data looks clean, nobody goes looking for the problem.
Most safety coordinators assume that a high BBS observation rate with flat incident numbers means the program is working fine and something else is driving the injuries. The real problem is usually the opposite: the observation program is producing noise, not signal, and leadership is using that noise to conclude the culture is healthier than it is.
When a plant runs rapid hiring cycles — say, adding 30 production associates over 90 days to meet a new contract — the people doing observations are watching a workforce that hasn't been trained to the same behavioral baseline. A new press operator on day 11 who hasn't yet internalized the lockout/tagout procedure for a specific machine model isn't going to exhibit the at-risk behaviors a trained observer knows to look for. The at-risk behavior is in the knowledge gap, not in the observable action. BBS doesn't catch that. The observation card comes back clean. The next week, that operator bypasses an energy isolation point because nobody showed them why it matters — just that they had to do it.
That's not a behavior-based safety failure in the traditional sense. It's a Management of Change failure that your BBS data masked.
Subcontractor crews introduce a structural problem that goes beyond training. A general contractor running welding and grinding work in your facility during a press line upgrade brings their own safety culture — or lack of one — onto your floor. Your BBS observers, if they're even watching the subcontractor crews at all, don't have the relationship or the authority to conduct a meaningful observation. They hand the sub's foreman a card and mark it observed. The foreman signs it and gets back to work.
The data says: observed, safe. Reality: a subcontractor grinder with no hot work permit training just created ignition risk twelve feet from your paint line.
Under OSHA's multi-employer worksite doctrine, your company can be cited as a controlling employer even when the at-risk behavior was performed by a sub. A serious violation runs up to $16,131 per instance. A willful or repeat citation — for example, if a prior inspection found the same hot work permit gap — runs up to $161,323. The observation card that said "safe" won't help you at that hearing.
Pull your BBS data and cross-reference it against your hiring and subcontractor activity over the same period. Look for these patterns:
The fix isn't more observation cards. It's requalifying your observers, establishing explicit protocols for observing new hires and subcontractor personnel within the first 30 days, and linking BBS findings directly to your onboarding and MOC processes.
A BBS program is a leading indicator only when it's measuring real behavior in real context. When turnover and subcontractor layering break the observer-observed relationship, the program starts producing a lagging echo — it reflects the culture you used to have, not the one currently on the floor. Your TRIR is flat or rising because the leading indicator stopped leading anything.
The Campbell Institute's research on leading indicators consistently points to one root issue: organizations measure the activity of a safety program (how many observations were completed) rather than the effectiveness of it (did behavior actually change). In a stable workforce, those two things correlate well enough that the distinction doesn't surface. In a high-turnover or subcontractor-heavy environment, they diverge — and the incident rate tells you the truth before the program data does.
OSHA's worker rights framework still holds the employer accountable regardless of who's on-site or how long they've been there. The observation card won't transfer that liability.
In most manufacturing plants running high turnover or subcontractor crews, the observation program is completing cards but not capturing real behavior. Observers are checking boxes with workers they don't know well enough to identify at-risk behaviors accurately. The metric reflects activity, not effectiveness.
Subcontractor personnel often fall outside the normal observer-observed relationship that makes BBS work. Observations of sub crews tend to be superficial because the observer lacks job-specific knowledge and authority. The result is data that shows "safe" while actual at-risk behavior goes uncaptured.
Under the multi-employer worksite doctrine, your facility can be cited as a controlling employer even when a subcontractor employee performed the at-risk behavior. Serious violations run up to $16,131; willful or repeat violations up to $161,323 per instance.
Audit the quality of observations, not just the count. Stratify your data by workforce tenure and work area. Requalify observers on what to look for with new hires. Build explicit observation protocols for subcontractor crews. Then connect your BBS findings to corrective action close-outs and track those separately.
The analysis above takes real time — time most safety coordinators in 50–300 person plants don't have because they're managing OSHA logs, training records, and written program updates instead. If the administrative load is the reason this work isn't happening, that's a solvable problem. Talk to EHS, Inc. — we handle the compliance paperwork so you can do the actual safety work.
Aaron West
Founder, EHS, Inc. — 18+ years in EHS compliance and contractor safety
Aaron West has spent over 18 years helping contractors and businesses navigate OSHA compliance, ISNetworld® certification, and workplace safety management. He founded EHS, Inc. to make enterprise-level EHS accessible to companies of all sizes — serving contractors and businesses nationwide — without long-term contracts or enterprise overhead.
Our team handles the complexity so you can focus on running your business. No long-term contracts, no learning curve.
Talk to EHSBBS programs count behaviors. HOP learning teams find the conditions that make those behaviors inevitable — and that's where your leading indicators actually live.
In upstream drilling and well servicing, the citation isn't for the hazard — it's for the missing verification record.
Your EMR tells you what already went wrong. Here's what data-driven safety pros in manufacturing track before the next incident.
Framework to achieve zero incidents
Stop hitting paywalls
54 topics in English & Spanish