The moment you operate a second location, safety changes character. You are no longer managing hazards; you are managing variation. And variation carries a specific regulatory cost: a hazard cited at one site and left uncorrected at another is the standard fact pattern for a repeat citation, which carries dramatically higher penalties.
The standard / implementation split
The structure that works is a clean division between what corporate mandates and what the site decides. Get this line wrong in either direction and the system fails — too centralized and sites ignore procedures that don't match their equipment; too decentralized and you have five programs and no ability to audit.
| Corporate owns (the what) | Site owns (the how) |
|---|---|
| Minimum standards and required programs | Machine-specific procedures for their equipment |
| Incident classification and reporting rules | Inspection scheduling and route |
| Training requirements by role | Training delivery and scheduling |
| Metrics definitions and reporting cadence | Local corrective actions and owners |
| Audit protocol and frequency | Day-to-day hazard control |
| Contractor qualification requirements | On-site contractor orientation |
The cross-site corrective action rule
This is the single highest-value multi-site control and it is one sentence: any finding at one site is evaluated at every site within 30 days. It applies to citations, serious near misses, incident root causes, and audit findings.
Without a cross-site propagation rule, every site has to learn every lesson by having its own injury. With one, the organization learns once.
Internal audits that aren't theater
- Use one protocol so results are comparable across sites — a shared scored checklist, not free-form observations.
- Have sites audit each other. Peer audits find more and transfer practices in both directions.
- Score consistently and publish the comparison. Visibility does more than mandates.
- Track finding closure rate as the metric, not finding count — a site reporting zero findings is not the best site, it's the least honest one.
Metrics for a multi-site view
- Normalize everything per 200,000 hours worked so sites of different sizes are comparable.
- Report leading indicators alongside lagging: inspection completion, near-miss rate, corrective-action on-time closure, training compliance.
- Watch the near-miss rate specifically — a site with a suspiciously low near-miss rate and average injuries has a reporting problem, not a safety advantage.
- Review the full scorecard monthly at the executive level with site leaders present.
Where fractional support fits
Multi-site organizations frequently over-hire — a safety manager at each location — before they have a corporate system for those managers to execute. The more efficient sequence is to establish the standard, the audit protocol, and the metrics framework first, with fractional senior leadership owning that layer, and then staff sites according to actual hazard load rather than symmetry.
Next step
Safety Exposure Score
Eight questions, two minutes. See where your program has real gaps.
Get Your Exposure Score