What to Report and to Whom

A behaviour system generates enormous quantities of countable material. Six numbers say whether anything is working.

The six

False trigger rate from clip review.

Time from event to coaching conversation.

Near-miss reports per hundred thousand miles.

Disputes raised, upheld and rejected.

Structural findings identified and acted on.

Collisions and severity, by type, over years.

Five of the six describe the programme rather than the people. That is deliberate.

What each says when it moves

False triggers rising: thresholds drifting, or new vehicle types added without retuning.

Time to coaching lengthening: the reviewer is overloaded, which is a resourcing finding.

Near-miss reports falling: cooperation is being lost, or an incentive is suppressing them.

Disputes at zero: the route is not believed.

Structural findings identified but not acted on: the programme is one-directional, and drivers can see that.

What not to report

A driver league table.

Depot rankings on event rates, which rank route types.

Individual monthly scores, which move on chance at that sample size.

Anything comparing drivers across vehicle generations, where assistance systems differ.

Who gets what

The programme owner: all six, monthly.

Depot managers: their own structural findings and coaching activity, not a ranking of their drivers.

Senior management: collisions, severity, and what was changed as a result.

Drivers: the false trigger rate, the structural findings, and their own data.

That fourth one is unusual and is the point — publishing the false trigger rate says plainly that the system is fallible, which is what makes people engage with it rather than resent it.

The annual view

Did collisions and severity move, over a long enough period?

Were the structural findings acted on?

Did near-miss reporting hold or rise?

Can the score still be explained to the person it describes?

And one honest statement about what the programme has not demonstrated, because a report of unbroken success is read once.

Publish the false trigger rate

To drivers, not only upward.

It says plainly that the system is fallible, which everyone already knows and nobody in management usually admits.

And it invites the corrections that improve it: the pothole, the map error, the loose device.

Very few programmes do this, and it is the cheapest credibility available here.

A practical configuration prompt

During configuration, use accounting time records for fleet reporting to prompt questions about purpose, ownership and output. Treat the page as a starting point and document each assumption.

Independent reference

For a thematic point of reference, see the International Labour Organization. This established source is a useful independent reference for the issue discussed here.