Near-miss reporting is the one metric where a low number is bad news. Organisations that report plenty are usually the ones finding things early; organisations that report almost none are usually the ones where nobody bothers.
Most attempts to fix this start with a poster asking people to report more. That has never worked, because the reasons people do not report are rational.
Why people do not report
Ask on site and the answers are consistent, and none of them are apathy:
It takes too long, and the form is on a computer in the site office
Nothing visible happened the last three times somebody did it
It felt like dobbing in a workmate
It was my own near miss and I did not want the conversation
I was not sure it counted
The two that actually matter
Friction and feedback. Everything else is downstream of those.
If reporting takes more than about a minute, on the device already in their hand, at the place it happened, it will not happen. And if reports disappear into a system that never visibly produces anything, reporting stops — not immediately, but within a couple of months, and permanently.
Close the loop in public
The highest-return change available is telling people what happened to the last report. Not a monthly statistic — the specific one: someone reported the step at the back of the shed, we replaced it, here it is.
This costs nothing and is the thing almost nobody does. One visible outcome does more for the reporting rate than any amount of encouragement, because it changes the implicit answer to "will anything come of this?"
Make it obvious what counts
A lot of under-reporting is genuine uncertainty. People default to reporting nothing when the threshold is unclear, so the threshold has to be stated plainly and generously: if it could have hurt someone and did not, that is one.
Photographs help more than text fields, because a photo of the thing takes five seconds and a written description of the thing takes two minutes and gets abandoned. Observations in HazNode are built around capture on a phone with a photo, for that reason rather than any other.
Where this meets your legal duty
Worth separating two things. Encouraging near-miss reporting is good practice, not a legal requirement. But some near misses are notifiable events, and those are not optional.
If an unplanned or uncontrolled event exposed someone to serious risk — a scaffold collapse, an unplanned explosion or fire, a substance escaping — it is notifiable whether or not anyone was hurt. Most near misses are not in that category. The ones that are need to go to WorkSafe as soon as possible, so the system that catches your near misses should be asking the notifiability question too. See notifiable events for what falls inside that.
Frequently asked questions
Is a low number of near-miss reports a good sign?
Almost never. It usually means reporting is too hard or that previous reports produced no visible result. Organisations with high reporting rates are generally the ones catching problems early.
How do you increase near-miss reporting?
Reduce friction and give feedback. Reporting should take under a minute on the device already in the person’s hand, and people need to see specific outcomes from previous reports. Closing the loop publicly on one report does more than any campaign.
Do near misses have to be reported to WorkSafe?
Some do. If an unplanned or uncontrolled event exposed someone to serious risk — for example a scaffold collapse, unplanned explosion or fire, or a substance escaping — it is a notifiable incident even if nobody was injured. Most everyday near misses are not notifiable.
Should near-miss reporting be anonymous?
It can help where people fear blame, but it has a cost: you cannot go back and ask the reporter what they saw, and you cannot tell them what you did about it. Many organisations get further by making reporting visibly consequence-free than by making it anonymous.
