When Safety Is an Afterthought

Safety Drift Syndrome

In high-velocity supply chain environments, safety is treated as a compliance overlay. A poster on a wall. A monthly drill. A line item in induction. Operations leaders assume that if no one has been seriously hurt yet, the system is working.

At a national last-mile network with 10,000+ workforce members across hubs, vehicles, and warehouses, the structural picture was different. Risky incidents — near misses, unsafe acts, equipment misuse, vehicle violations — were running unchecked. Most were never reported. The few that surfaced went into a complaint pile with no owner and no closure rhythm.

The symptoms were familiar to anyone who has scaled physical operations:

Underreporting of near misses. Frontline workers had no incentive to report. Reporting created investigation overhead and rarely led to change.

No data, only anecdotes. Leadership debated safety using stories, not patterns. Decisions were reactive to the loudest incident.

Safety owned by no one. HR thought it was Ops. Ops thought it was HR. Hub managers thought it was head office. Head office thought it was the hub.

Compliance theatre. Safety reviews focused on paperwork completeness, not behavioural drift on the floor.

Safety was not failing because people did not care. It was failing because no structure was holding it.

What the Surface Symptoms Were Hiding

A conventional safety lead would have run more trainings, printed more posters, and tightened the audit checklist. None of that addresses the underlying structural diseases.

Two diseases from the MetMov taxonomy were active simultaneously:

1. Process Drift. Safety SOPs existed on paper but had decayed in practice. The gap between the documented standard and the floor reality had widened year over year, and no one was measuring the drift.

1. Accountability Vacuum. No single role owned the safety outcome at the hub level. Without a named owner with thresholds and consequences, safety became everyone's general concern and no one's specific job.

A 5x gap between reported and actual incidents is not a training problem. It is a reporting architecture problem layered on top of an ownership problem.