Most multi-site companies default to one resident safety manager per plant, and most of them get worse safety performance for it than a leaner structure would produce. Not because the person is weak — they’re usually one of the best people in the building. Because a resident safety role absorbs accountability that belongs to the line, and the org design does that automatically, regardless of who fills the seat. The highest-performing multi-site operations I’ve seen run thin corporate EHS teams and thick line ownership. That’s not a coincidence.
Why does headcount feel like the answer?
Because it’s the most legible response to a real problem.
Something happened. The GM needs to demonstrate a response. A req for a safety manager is visible, budgetable, and defensible to a board or a carrier. Redesigning how supervisors run a pre-shift huddle is none of those things, even though it’s the thing that would actually move the number.
There’s a second driver, and it’s more honest: operating leaders are overloaded, and a safety manager makes the load go away. That’s not cynicism. It’s the predictable result of adding a role whose implicit job description is “handle the safety stuff.”
So the req gets filled, and the org quietly reorganizes itself around the new person in a way nobody intended.
How does a resident safety manager become a crutch?
Watch the sequence. It’s the same everywhere.
The safety manager starts on the floor doing real work — assessments, coaching, fixing conditions. They’re effective. Word spreads that safety questions go to them.
Within a quarter, they’re doing the incident investigations. Not facilitating them — doing them. The supervisor who was there gives a statement and returns to the line.
By month six they own the corrective action list, which means they’re chasing other departments’ commitments without any authority over those departments. They write the training, deliver the training, and track the training. They attend the audits. They answer the carrier’s questions.
By month twelve, the plant has a stable equilibrium: safety is a function, staffed, with a name on it. The line’s honest belief is that they support safety, and support is a very different thing from owning it.
None of this is the safety manager’s failure. Every step was them doing their job well. The design converts a capable professional into the organization’s accountability sink, and then measures them on results they don’t control. That’s why turnover in these roles is what it is, and why the second hire rarely performs better than the first.
The tell is simple. Ask a supervisor who owns safety on their line. If the answer names a person in the safety office, the design has already failed, and hiring another one deepens it.
Do lean structures actually perform better?
Look at the operations you’d consider world-class in this space — large manufacturers running multi-site or business-unit EHS coverage with a handful of senior resources. They exist, and the pattern is consistent: not more safety headcount, but more line ownership with senior EHS setting standards, building capability, and governing.
The direction of causation is worth being careful about. Lean EHS staffing doesn’t produce good performance. Good performance permits lean staffing — you can run thin when the line genuinely owns the work. But the reverse is also true and less obvious: heavy per-site staffing actively prevents the transition, because it removes the pressure that would otherwise force ownership onto the line.
You can’t get to the mature state while continuing to pay for the structure that keeps you out of it.
What does the maturity ladder say about staffing?
We use a five-stage EHS Maturity Ladder — Reactive, Compliant, Managed, Integrated, Self-Correcting — to diagnose where an operation actually sits. The staffing implication changes at each stage, and getting it backwards is expensive.
Reactive and Compliant. You genuinely need capacity. There’s a real backlog — programs to write, training to deliver, records to rebuild, conditions to fix. Buy hands for this. Contract support, a coordinator, a technician. Don’t buy a senior leader to work a backlog, and don’t confuse the backlog phase with a permanent headcount need.
Managed. This is the hinge. Leading indicators are being tracked, corrective actions close on schedule, and the question becomes who runs the cadence. If the answer is the safety manager, you stall here permanently — and this is exactly where most multi-site companies live. Getting past it requires moving cadence ownership to operations while EHS moves to standard-setting and governance. It is a deliberate act, and it will be uncomfortable for a quarter.
Integrated. Safety is inside the operating system — on the tiered board, in standard work, in capital review, in the operating review your leaders already attend. At this stage per-site senior EHS headcount is genuinely redundant. What you need is coverage for technical depth — industrial hygiene, PSM if you have covered processes, environmental permitting — and that depth is better bought as expertise than as a resident generalist.
Self-Correcting. The system finds and fixes its own drift. Rare. Worth naming so people know the ladder doesn’t end at Integrated.
The diagnostic failure I see most often: a company at Managed hiring for the Reactive problem, because that’s the hire they know how to make.
When do you genuinely need someone on site full-time?
I’d rather be specific than sell a model that doesn’t fit.
You need resident presence when the hazard profile demands daily technical judgment — PSM-covered processes, significant confined space or energy control volume, heavy construction with an active permit environment, or high-hazard startups and commissioning.
You need it at scale. A single site with well over a thousand employees across multiple shifts has enough sustained volume of legitimate work to fill a seat and then some.
You need it under enforcement or settlement obligations with monitoring requirements.
And you need it during a genuine crisis — post-fatality, or an operation in freefall where someone has to be in the building every day while the system gets rebuilt.
Outside those conditions, examine the assumption. Most single-site manufacturers under a few hundred employees, and most multi-site portfolios of small plants, are buying presence when what they need is direction plus capacity — and those are two different purchases, priced very differently.
What’s the alternative structure?
Three layers, and only one of them is senior.
Senior EHS leadership, fractional. Strategy, standards, governance cadence, leadership alignment, accountability for the roadmap. Part-time, because the strategic work is not forty hours a week even at a multi-site portfolio. This is the seat FractionalEHS occupies.
Execution capacity, sized to the actual work. Coordinators, technicians, or contract support for training delivery, inspections, records, and documentation. Cheaper than a manager and better matched to what the work actually is.
Line ownership, made explicit. Supervisors and plant managers own conditions, corrective actions, and the daily loop on their line. Not “supported by safety.” Owned, on the tiered board, with the same consequence structure as their other commitments.
That structure costs less than a resident manager per site and produces a different result, because it puts accountability where the authority already is.
The transition is the hard part, and I won’t pretend otherwise. There’s a quarter where line leaders are visibly uncomfortable and the safety function feels like it’s abandoning them. Executive air cover during that quarter is the entire difference between the redesign working and reverting.
What should you do before you post that req?
Ask three supervisors who owns safety on their line and listen to the exact wording.
Then take whatever work you were about to hire for and sort it into two columns: senior judgment, and volume. Look at the ratio. If it’s mostly volume, you’re about to overpay for a manager to do coordinator work. If it’s mostly judgment, you don’t need forty hours of it.
Then ask what your safety manager would be accountable for that they can’t actually control. If the list is long, the problem isn’t a vacancy.
