Metric dashboard design narrative
How to structure a dashboard that tells a story, not just a grid of numbers.
1. Start with the question
Every dashboard answers one question. Define it before you touch a chart. “Are we growing?” is different from “Where are users dropping off?” The question dictates metric selection, granularity, and layout priority.
2. Narrative hierarchy
Place the headline metric top-left — the single number that answers the core question. Below it, trend lines show direction. Supporting metrics fan out to the right and bottom. A viewer should read it like a newspaper: headline, lede, details.
3. Time as the spine
Anchor every section to a time axis. Current period vs. prior period. Rolling 7-day average. Year-over-year. Without a temporal anchor, numbers float without meaning. Show change direction with color — violet for positive, pink for negative.
4. Segment to reveal
A single aggregate hides the story. Break metrics by cohort, plan tier, geography, or acquisition channel. The dashboard should surface which segment drives the trend. Use small multiples — repeated mini-charts per segment — rather than one overloaded plot.
5. Thresholds and alerts
A metric without a threshold is noise. Define warning and critical bands for every KPI. When a metric crosses a band, the dashboard should surface it visually — a subtle border pulse or color shift — not bury it in a notification bell.
6. Action links
Every metric block should answer “what do I do with this?” Link to the relevant drill-down, alert configuration, or external tool. A dashboard is a launchpad, not a destination.
Pro tip
Before shipping, show your dashboard to someone outside your team. Ask them to narrate what they see. If their story doesn’t match your intended narrative, restructure.