Concept
1. A dashboard is a decision tool
A report answers "what happened?". A dashboard answers "is everything OK, and if not, where do I act?" — in under a minute. If nobody changes a decision after looking at it, it's decoration.
2. The dashboard brief (fill this first)
| Question | Example answer |
|---|---|
| Who is the main viewer? | Owner / Sales head |
| What decisions will they take with it? | which region/store needs attention, whether to push a product, discount approval |
| Which 3–5 questions must it answer? | Are we on target? Which region is behind? Which products drive revenue? Who are the top/bottom salespeople? |
| How often will they look? | every Monday morning |
| Where? | laptop screen in a meeting; sometimes printed A4 |
| How much time will they spend? | 1–2 minutes |
| Comparison they care about | vs target, vs last year |
| Level of detail | region → store; not individual orders |
Every chart, card and slicer must trace back to one of the questions. If it doesn't, leave it out.
3. Same data, three audiences
| Audience | Needs | Dashboard style |
|---|---|---|
| Owner / CEO | "Are we OK?" — few numbers, targets, trend | 4–5 KPI cards, one trend, one exceptions list |
| Sales manager | where to act this week | region/store comparisons, leaderboard, slicers |
| Analyst | why it happened | many filters, detail tables, drill-down |
Trying to serve all three in one sheet usually serves none. Build the owner view first; link to detail sheets for the rest.
4. Strategic, tactical, operational
- Strategic (monthly/quarterly): trends, targets, YoY — slow-moving.
- Tactical (weekly): comparisons across regions, products, people.
- Operational (daily/live): today's orders, stock alerts, pending tasks.
Decide which one you're building — it sets the refresh frequency and the level of detail.
5. Interview, then sketch
Spend 15 minutes asking the viewer: "What do you check first? What makes you worried? What do you do when a number is bad?" Then sketch the layout on paper (boxes and labels, no colours) and show it before building. Changing a sketch costs a minute; changing a finished dashboard costs a day.
Common mistakes
Starting from "what data do we have?" and charting everything. One dashboard for every audience. No comparison (a number without target or last year means nothing).