Big companies buy the wrong software all the time. They watch a smooth demo, sign a three year deal, then find out the tool cannot handle their sites, their rules, or their people. Picking enterprise incident management systems takes more care than picking a tool for one small team. It has to work across many sites and many roles. Australia had 146,700 serious workers’ compensation claims in 2023-24. At that scale, a small gap in software becomes a big gap in safety, and the cost of a bad choice repeats every single day. Here is what I would check first.
Which features matter most?
| Feature | Why it matters | Red flag |
| Mobile reporting | Workers log events on site | Needs a desk computer |
| Role based access | Right people see right data | Everyone sees everything |
| Custom forms | Each site has its own risks | Vendor charges for every change |
| Automatic alerts | Managers hear fast | Only a daily email |
| Dashboards | Leaders see trends across sites | Exports only, no charts |
| Audit trail | Shows who did what, and when | Records can be edited quietly |
Six rows, and to me they are all deal breakers. If I had to pick one, it is the audit trail. A system that lets records change without a trace is a liability, not a tool. Ask to see the log for a record that was changed. If they cannot show one, walk away.
Can it grow with your business?
Ask about limits first. How many users? How many sites? What happens when you open in a new state or country, where the rules differ? Test it with your own messy data, not the demo data. A system that works for 50 people can crawl at 5,000. Ask for one reference at your scale. Demos are made to look good. Real incidents are not.
How well does it connect to your other tools?
An incident system on its own is an island. It should talk to your HR system, your training records, and your risk register. If it cannot, someone will copy data by hand, and copying by hand is where mistakes breed. Ask to see a working link, not a slide about one. Also check the boring parts, like single sign on and how updates roll out.
What should you ask every vendor?
- Where is our data stored, and is it in Australia?
- Who can see it, and is that logged?
- What does it cost when we add users or sites?
- Can we take all our data with us if we leave?
- How long does setup take? Show real examples.
- Can we talk to a customer of our size?
Push hard on that export question. Getting locked in is the classic regret. Ask for the answers in writing.
Who will actually use it?
Workers, supervisors, managers, and safety teams all use it in different ways. Ask each group to test it. A supervisor in gloves, with a bad signal, should finish a report in under three minutes. If your frontline staff hate the tool, they will stop reporting, and then your fancy dashboard shows you nothing.
Should you care about mental health reporting?
Yes. Claims for mental health conditions are now 12% of serious claims. That is up 14.7% in a year and up 161% in ten years. Those claims also keep people away from work almost five times longer than other injuries do. Yet many systems only have boxes for broken bones. Check that yours can log bullying, violence, and heavy workloads. Most Australian jurisdictions now expect employers to manage these risks. Better to find out now than after a claim.
