Why the data matters
Every serious incident has near misses that preceded it — the same hazard, the same conditions, without the unlucky timing that turned it into an actual injury. A company that only tracks actual incidents is only seeing the failures that got expensive. Near miss data shows the same underlying risks before they cost anything, while there's still time to fix the conditions that caused them.
Why workers don't report them
A few consistent reasons show up across construction crews:
- Nothing actually happened, so it doesn't feel worth flagging
- Reporting feels like admitting a mistake, even when the hazard wasn't the worker's fault
- The reporting process itself is friction — a form to track down, a supervisor to find, paperwork after an already long shift
- No visible follow-up on past reports makes the next one feel pointless
None of these are really about the value of the data — they're about the cost of reporting it, both real and perceived.
Counterintuitively, a crew with more near miss reports isn't necessarily a more dangerous crew — it's often a crew with a stronger reporting culture. Low near miss numbers combined with occasional real incidents is a warning sign, not a good one.
What actually changes reporting behavior
Making the report itself fast — genuinely under a minute, from a phone, at the moment the worker is already thinking about what just happened — removes most of the friction. So does visible follow-up: if workers see that a reported hazard actually gets addressed, the next report comes easier. A system where reporting takes as long as any other daily task, rather than being treated as a separate, heavier process, changes the actual behavior far more than a policy reminding people to report more.