Information & decisions · JNAGA perspective
Why does a dashboard show numbers nobody trusts?
A dashboard can make disagreement visible. It cannot settle what a measure means by drawing it more attractively.
Published 25 September 2026
Start with the management question
What action should a person take after seeing the measure? If the answer is unclear, the dashboard may be reporting activity rather than supporting a decision. Name the event being counted, the period and the exclusions in ordinary language.
A measure can be technically correct and still misleading if it hides reopened cases, partial work or a changed process. Compare several underlying records with the displayed number before redesigning the chart.
Trace where the number changes
Follow the measure from source record through any export, integration or manual adjustment. Note when each step updates and who can repair an error. This often reveals that the problem is not the visualisation but the inconsistent record beneath it.
Make uncertainty visible. A provisional or incomplete count can still be useful if its limits are clear and the decision does not demand false precision.
Build a correction route
Give users a way to question a figure and understand the answer. Assign ownership for the definition and the data, then revisit both when operations change. Trust grows through transparent correction, not a more elaborate dashboard.
Imagine a weekly dashboard says twelve requests were completed while the operations team remembers ten. Two requests were reopened after an initial closure.
Both figures may follow different rules. Decide whether the management question concerns first closures, final resolutions or work still needing attention, then show that definition and the exceptions beside the number.