Safety diligence on a manufacturing acquisition typically consists of three artifacts: the TRIR trend, the OSHA 300 logs, and a compliance rep in the purchase agreement. All three are backward-looking, all three are curated by the seller, and none of them measures the thing that will actually determine your exposure through the hold period: whether safety at this company is a system or a person. The first 90 days after close are when you find out — and when it’s cheapest to fix.
Watch: Just Acquired a Manufacturer? Here’s What Diligence Didn’t Tell You
What does diligence actually see?
The data room shows outcomes, not conditions. A clean five-year TRIR trend can coexist with an unguarded palletizer, a lockout program that lives in one maintenance lead’s head, and a Dust Hazard Analysis citing a standard that no longer exists. I’ve written about why the recordable rate flatters latent severity exposure — in an acquisition context, that flattery is worse, because the seller had every incentive to polish the number and none to surface the deferred maintenance behind it.
The compliance rep doesn’t save you either. Reps and warranties transfer legal recourse, not operational risk. When the serious injury happens in month seven of your hold, the indemnity conversation is a sideshow next to the OSHA response, the insurability conversation, and the management distraction — all of which are now yours regardless of what the seller warranted.
Where does the real exposure hide?
Four places, reliably.
The maintenance backlog. Guarding, energy isolation hardware, mobile equipment segregation — the capital items a seller defers in the two years before a sale because they don’t show up in EBITDA. Pull the work order history, not the policy binder.
The single point of failure. Many mid-size manufacturers run safety through one person, and acquisitions are exactly when that person leaves — new owners, new uncertainty, better offers. If the program is that person, your program has a notice period.
The insurance trajectory. The experience mod you inherited was priced on the seller’s history; the claims developing right now are priced into yours. A mod that’s about to move changes your workers’ comp line for three years — quantify it in the first month, not at renewal.
The maturity gap between what leadership believes and what the floor runs. On the EHS Maturity Ladder — Reactive, Compliant, Managed, Integrated, Self-Correcting — most acquired manufacturers sit at Compliant while their new deck says Managed. The tell is simple: ask who owns the question “what changed in our regulatory environment this year,” and watch whether anyone answers by name.
What should the first 90 days look like?
Not a 200-item audit. Four moves that fit inside integration without competing with it.
Days 1–30: walk the floor with fresh eyes and rank by energy. High-energy exposures first — falls, mobile equipment, hazardous energy, confined spaces. You’re not building a compliance list; you’re building the severity map diligence never drew.
Days 15–45: pull the three histories. Work orders on safety-critical equipment, the loss runs and open claims behind the mod, and the last two carrier loss control reports with their open recommendations. Together they tell you what the company actually spent, deferred, and ignored.
Days 30–60: name the owner and test the dependency. Identify who actually runs safety and what leaves if they do. If the answer scares you, that’s your first hire-or-bridge decision — and it’s the moment operator-grade fractional leadership earns its fee: senior judgment installed at the top of the org while integration is still fluid, without adding permanent headcount before you understand the operation.
Days 60–90: put safety into the value creation plan with a number. Guarding capex, mod trajectory, and the two or three systems that must exist by month twelve. Safety exposure that isn’t in the plan gets managed by whoever has spare time, which is no one — and it stays invisible until it reprices the exit.
“We priced it into the model” — did you?
The steelman objection: sophisticated buyers discount for operational roughness, so the exposure is already in the price. Sometimes true for the capex line. Almost never true for the tail risk. A fatality in the portfolio doesn’t cost what the model discounted — it costs the human toll first, then litigation, insurability, regulator attention across the platform, management bandwidth for a year, and a story every future buyer will ask about. You cannot price that in. You can only build the system that makes it unlikely, and the window where that’s cheapest is the one you’re in right now.
Key takeaways
- Diligence sees curated outcomes; ownership inherits conditions. TRIR trends and OSHA logs are backward-looking and seller-polished.
- The exposure hides in four places: the maintenance backlog, the single-person program, the developing mod, and the gap between claimed and actual maturity.
- Reps transfer recourse, not risk. The incident in month seven is yours no matter what the seller warranted.
- The first 90 days are a severity-mapping exercise, not a compliance audit — energy first, histories second, ownership third, plan fourth.
- Tail risk can’t be priced in, only engineered down — and post-close is the cheapest moment you’ll ever have to do it.
