Dashboard Data Refresh Audit (QA)

Dashboard Data Refresh Audit (QA)

1. 01

   Freeze the metric definitions before testing any refresh

   Pull the signed metric dictionary from onboarding and confirm every KPI on the dashboard maps to exactly one definition, owner, and source system.

2. 02

   Reconcile each dashboard figure against its system of record

   Pick three high-stakes numbers, trace them back to the source query, and document any variance above 1 percent with the reason it exists.

3. 03

   Test the refresh cadence against the client's decision rhythm

   A daily refresh on a metric the client reviews monthly adds noise; a weekly refresh on a metric tied to a Monday standup arrives too late.

4. 04

   Verify stale-data handling is visible, not silent

   Confirm the dashboard shows a last-updated timestamp and a clear warning state when a source fails, rather than displaying yesterday's number as current.

5. 05

   Check access permissions against the actual audience list

   Board-level viewers, client department heads, and agency delivery leads often need different slices; confirm no one sees a metric they were not cleared for.

6. 06

   Run one full failure drill on the weakest data source

   Disconnect the least reliable feed, confirm the dashboard degrades gracefully, and record how long the team took to notice and respond.

7. 07

   Log every open data-quality issue with an owner and a resolution date

   An unresolved variance that sits in a shared doc for three weeks becomes a credibility problem the next time a client questions a number.