How to measure wellbeing without surveilling staff

If your dashboard can show Thandi’s mood on Tuesday, it is the wrong dashboard. Trust will collapse the first time someone notices, and they will notice.

Useful numbers are boring. What share of invited people opened a lesson. What share finished the programme they were assigned. Whether check-in participation rose or fell. All of it for a group large enough to hide the person.

Set the floor at five, and mean it. A filter that drills into a branch of three people should show nothing, not a clever average.

Do not mix surveys with discipline. The moment a low score can affect a bonus, people will lie, and your trend becomes fiction.

Read the report once a month, write two actions, and tell the team what you changed. Measurement that never changes the work is just another form.