Incident reporting software: getting the report in the first place

Incident reporting software has one job before all the others: getting the event recorded at all. Triage, investigation, trend analysis and statutory notification are downstream of a person deciding it is worth their time to tell you something went wrong. Most systems are designed as though that decision is guaranteed. It is not — every extra field, login and approval step quietly reduces the number of events you ever hear about.

What incident reporting software must handle

• Capture in under a minute, from a phone, offline, with a photo and free text as the minimum viable report.
• Anonymous or confidential reporting, not routed by default through the person's own manager.
• Triage by severity, with a defined response time for each level.
• Investigation depth proportionate to consequence — not the same form for a grazed hand and a fatality.
• Actions with owners, dates and verified closure, tracked separately from the investigation write-up.
• Statutory notification where a jurisdiction requires it, with its own clock.
• Trend reporting by location, task and cause, which is the only way the data pays you back.
• Linking related events, so a recurring low-severity pattern is visible as one problem.

Reporting rate is the metric

A falling incident count is ambiguous and usually misread as improvement. The number that tells you whether the system works is how many low-severity events and near misses arrive per period: that volume should be high and rising as trust grows. Two lost-time injuries and zero near misses is not a safer picture than four hundred near misses and none — it is a picture of an organisation that will find out about its problems through consequences rather than through reports.

Close the loop visibly. The single strongest predictor of whether somebody reports a second time is whether anything observably happened after the first. Publish what changed, even when the answer is that you assessed it and decided no change was needed.

Making it work

1. Cut the capture form to what you genuinely need at the moment of reporting; ask the rest during triage.
2. Set and publish the response time per severity level, then measure yourself against it.
3. Separate the corrective action from the immediate fix, and verify closure independently.
4. Report the reporting rate to management alongside the incident count, and explain why.
5. Review recurring low-severity events quarterly — that is where the next serious one is signposted.
6. Never let a report be closed by the person whose decision is under examination.

Ettex Records covers the register without a platform: one row per event with date, location, severity and description, photos attached, triage and investigation notes, and an action log with owners, due dates and verification. Filtering by severity and by month gives the trend view, and the same structure supports an incident report without maintaining two systems.