A dashboard can refresh on screen without receiving new data. A “last updated” label may refer to the page, the report calculation, or the source extraction. Those are different events, and the distinction matters when someone is making a time-sensitive decision.
Identify the useful timestamp
Show when the relevant source data was successfully refreshed. If several sources update separately, do not imply that one timestamp applies equally to all of them. A clear note can say which part of the view is delayed.
Choose a freshness expectation that fits the decision. A monthly planning report does not need the same update cadence as a daily scheduling queue. More frequent refreshes add operating cost and complexity without always improving the decision.
Make stale data visible
Imagine a hypothetical booking dashboard whose schedule data is current but staffing availability has not synchronized. A combined capacity number may look precise while using an outdated input. The report should flag the affected measure and direct the operator to the authoritative source if needed.
Avoid replacing a failed update with zero. Missing data and no activity are different states. Preserve the last successful value with an explicit stale label when that is the appropriate design.
Define the response to a delay
Assign an owner for failed refreshes and identify when a delay warrants attention. A freshness warning should explain what the viewer can safely use and what requires checking elsewhere. Repeated alerts without an action route quickly become background noise.
Test an unavailable source and a partially completed update. Confirm that the dashboard does not silently combine incompatible reporting periods.
Start by tracing one important number from its source timestamp to its display. The useful result is a viewer who knows how current the evidence is before acting on it, rather than discovering the delay after the decision.

