Data Dashboard
Build a dashboard as a decision interface, not a collection of charts.
- Define audience, decisions, owner, cadence, data freshness, metric contracts, and action thresholds.
- Limit the primary view to the few measures that change decisions; place diagnostic detail behind drill-downs.
- Specify filters, segmentation, comparison baselines, empty/error states, and point-in-time semantics.
- Validate each displayed measure against the source/query and show data freshness/known gaps.
- Review usability, accessibility, and performance before publishing.
Do not silently create external dashboards, connect data sources, or send alerts. Treat threshold colors as decision cues backed by a named owner.