EHS Software Selection Checklist

1. Write the field task in the user's terms

Describe who reports, what information they need and where the task takes place. Test the intended device, language, accessibility needs and shared-device arrangements. If connectivity is a constraint, specify which tasks must remain available, how pending work is shown and how records are reconciled after reconnection. A mobile app label does not answer these questions.

2. Follow the record through closure

Demonstrate initial review, classification, investigation, action assignment and verification using the workflow you expect to configure. Ask who can reopen an item and what happens when ownership changes. Check the evidence behind dashboard counts so that submitted, overdue, completed and verified records mean the same thing to your team and the vendor.

3. Check sensitive access and content ownership

Use synthetic data to test whether a reporter, supervisor, investigator and administrator see the right fields and records. Define which information should be collected and retained. Where regulatory registers or reporting templates matter, identify the jurisdiction, content provider, version, update process and person responsible for checking applicability. A template library is not evidence of compliance.

4. Distinguish configuration from further work

Record whether each demonstrated feature is included, configured, supplied by a partner or still proposed. Ask who maintains forms and workflows after launch and which changes require paid assistance. For an interface, identify the exact record transfer, responsible system and recovery process. Include migration, training and ongoing administration in the cost comparison.

5. Record the evidence before calculating a score

For each priority task, record “demonstrated”, “partly demonstrated” or “not tested”, along with the observed result. Keep critical failures separate from weighted preferences so a high overall score cannot hide an essential gap. Agree which open items need a follow-up demonstration and which require a pilot with representative users.

6. Check how records are kept and corrected

For United States sites, OSHA's recordkeeping rule requires the 300 Log, the annual summary and the 301 incident report to be kept for five years following the end of the calendar year they cover. It also requires the 300 Log to be updated when a case changes: if the description or outcome changes, the original entry is removed or lined out and the new information entered. In a demonstration, reclassify a sample case from restricted work to days away. Check whether the earlier entry stays visible in the history, whether the log view shows the change, and whether you can set retention to your own period and still delete when it ends.

OSHA: 29 CFR 1904.33, retention and updating. The text of the US rule for establishments covered by it. Other jurisdictions set their own retention periods. Checked 3 October 2026.

7. Ask which report each field feeds

An incident form often feeds several outputs: an internal investigation, a statutory log and sometimes an agency submission. OSHA's recordkeeping page says that many employers with more than 10 employees must record work-related injuries and illnesses on Forms 300, 300A and 301 or equivalents, that some establishments submit data electronically once a year, and that every employer must notify OSHA within 8 hours of a work-related death and within 24 hours of an in-patient hospitalisation, amputation or loss of an eye. Ask the vendor to show which of your fields feed each output, who decides whether a case is recordable, and whether the notification clock is captured or only the final form.

OSHA: recordkeeping. A summary of US requirements, with industry exemptions. It does not say how any software handles them. Checked 3 October 2026.

8. Test your own jurisdiction's definitions with two events

In Great Britain the equivalent duties sit under RIDDOR, the Reporting of Injuries, Diseases and Dangerous Occurrences Regulations, and HSE's page lists a record of reportable injuries, diseases and dangerous occurrences among the things to keep. A tool built around United States forms may have no field for a dangerous occurrence, and a tool built for Great Britain may not classify a case in OSHA terms. Create two sample events, one reportable under the rules that apply to you and one that is only an internal near miss. See how each is classified, by whom, and whether the classification can be challenged and the reason recorded.

HSE: RIDDOR. HSE's entry page for reporting in Great Britain. The two-event test is our proposal. Checked 3 October 2026.

9. Let the people who report try it

HSE presents consulting workers as four stages: prepare, plan, consult and improve. A tool selection can be run as one such consultation. Ask a few reporters, supervisors and a safety representative to complete the field task and tell you what slowed them down, and put their comments in the checklist rows next to the observed result. A tool that suits the administrator but frustrates the people who file reports yields fewer and thinner records, and no dashboard will show the difference.

HSE: worker involvement. HSE guidance on involving and consulting workers in Great Britain. It does not address software selection; applying it that way is our reading. Checked 3 October 2026.