Take last month's safe manhours. If you asked three people, would you get the same number?
Usually not — and usually for defensible reasons. One counts contractors, one does not. One counts a six-day week, one assumes five. One counts supervisory hours on site, one counts only direct labour. Every version is calculated correctly. They are answers to three different questions that have all been given the same name.
The consequence is not inaccuracy. It is that the meeting stops being about the operation. A manager presents a rate, somebody challenges the denominator, and the next twenty minutes go on reconciling a number instead of deciding what to do about it. People learn that bringing a figure invites an argument, so the figures get softer, and the review becomes a formality.
The fix is boring and it is not a dashboard. It is a written definition for each number that matters: what is counted, what is excluded, who produces it, and from which source. One page. The moment that exists, two of the three versions become wrong in a way anybody can check, and the argument has somewhere to end.
Only then is a dashboard worth building. A dashboard over undefined numbers does not resolve the disagreement — it renders it faster, in colour, and gives it the authority of a screen.
Your safe manhours for last month — if you asked three people, would you get the same number?