Nexus Distribution: falling near-miss reports and a £40,000 choice

Why the numbers are a warning, not an improvement

• Near-miss reporting is a leading indicator of reporting culture, not of risk. A fall from 34 to 6 in the same quarter that a racking strike went unreported points to suppression, not to safer working.
• The cause is visible in the history: an operator who reported a dropped pallet was moved off the truck team. Staff learn from what happens to colleagues faster than from posters, and supervisors still retelling the story keeps that lesson alive.
• The organisation is measuring itself on a lagging, blame-attracting number (reportable injuries). When a metric can only get worse by being honest, people manage the metric - by not reporting, by reclassifying injuries as first aid, or by reporting damage only when an inspection finds it.
• Agency staff on weekly contracts have the least security and the most to lose from reporting; a 20-minute video induction gives them no relationship with a supervisor who might back them.

What the adviser should do

1. Treat the near-miss drop as a finding to investigate, not a result to celebrate: interview a sample of drivers and pickers, run an anonymous survey, and compare near-miss numbers against racking damage found at inspection, FLT telematics/impact data and absence records.
2. Publicly close the loop on the old case - explain what happened and what has changed - so the story stops being told as a warning against reporting.
3. Rebalance the scorecard toward leading measures the site controls: inspections completed, corrective actions closed on time, damage found per inspection, supervisor safety conversations, and reporting volume with a quality review rather than a target.

The judgement being tested is that a falling incident count is ambiguous evidence, and must be triangulated before it is believed.

Reasoning first, spend second

The two options differ in reliability, not just in cost. Racking barriers are an engineering control: once installed and inspected, they work whether or not anyone is motivated, tired, new, or agency. Racking strikes are a high-consequence, low-frequency event - a collapsed bay can kill a picker - and the six worst aisles are already showing damage, which is direct evidence of both likelihood and current control failure.

A behaviour-based safety programme depends on trust. Observers must report what they see without fear, and workers must believe reports lead to fixes rather than discipline. On this site a reporter was moved off the truck team six months ago. Launching BBS into that culture risks confirming the belief that observation is surveillance, and it takes months before effects appear.

Recommendation

• Allocate the majority to physical protection: barriers or sacrificial guards on the highest-risk uprights, prioritised by damage history and proximity to pick faces and pedestrian routes, plus end-of-aisle protection.
• Retain a modest slice for the culture work that costs little: reinstating a visible 'you said, we did' feedback loop, supervisor coaching on receiving reports, and a revised induction for agency staff that includes a named person to report to.
• Defer the external BBS contract until reporting volume has recovered for at least a quarter - BBS is an accelerator, not a repair tool.
• Record the decision and reasoning, and set a review date.

Justify to the operations director in risk terms, not preference: the chosen control does not depend on human behaviour, addresses the most severe credible outcome, and buys time to rebuild reporting before asking staff to be observed.

What a zero target actually drives

A target of zero reportable incidents is a target on an outcome that is partly random and largely invisible. Staff and managers cannot control whether an injury happens this year in any reliable way, but they can control whether it is reported and how it is classified. The predictable consequences are:

• under-reporting, especially by the least secure workers - agency staff on weekly contracts;
• pressure on first-aiders and supervisors to record injuries as first aid, or to persuade staff to take annual leave rather than be absent;
• accidents pushed off-site or attributed to the individual ('didn't follow technique'), so the organisation loses the learning;
• a false 'excellent record' that collapses at the first serious incident, often with a much larger enforcement and reputational cost than the original problem.

There is also a moral point the director should hear: a zero target tells an injured worker that their injury was someone's failure of will, which encourages concealment and delays treatment.

What to propose instead

A balanced scorecard with a clear split:

• Leading and controllable: racking inspections completed on time; percentage of reported damage rectified within a set period; near-miss reports per 100 staff with a quality audit of a sample; closed-out corrective actions; supervisors trained in report handling; induction completion for agency staff before first shift.
• Lagging, for monitoring only, without a target: reportable injuries, over-7-day injuries, and - crucially - the ratio of near-misses to injuries, which should rise as reporting improves.

Challenge the director on how the target would be achieved, and make the case that a rising near-miss count with fast fixes is a more defensible position in front of an enforcing authority than a flat line of zero reports.