Dashboard & KPI Designer

When to Use

• When building a new operational or executive dashboard
• When KPIs exist but lack clear targets, thresholds, or visualization
• When stakeholders ask "what should we measure?"
• When consolidating scattered metrics into a single source of truth

Procedure

Step 1 — Objectives Alignment

Clarify what decisions the dashboard must support:

• Who is the primary audience? (executive, manager, analyst, operator)
• What decisions do they make with this data?
• What cadence? (real-time, daily, weekly, monthly)

Step 2 — KPI Selection

For each objective, define KPIs using the SMART-KPI framework:

KPI	Type	Target	Threshold (Green/Amber/Red)	Data Source
[name]	Leading / Lagging	[value]	G: ≥X / A: Y–X / R: <Y	[source]

Rules:

• Max 7 KPIs per dashboard view
• Every lagging indicator must have ≥1 leading indicator
• Every KPI must have a named owner

Step 3 — Visualization Selection

Data Pattern	Recommended Chart	Avoid
Trend over time	Line chart	Pie chart
Part-to-whole	Stacked bar, treemap	3D charts
Comparison	Bar chart, bullet chart	Radar chart (>7 axes)
Single value vs target	Gauge, big number + trend arrow	Table
Distribution	Histogram, box plot	Line chart
Relationship	Scatter plot	Stacked area

Step 4 — Layout & Hierarchy

Build the dashboard with visual hierarchy:

1. Top row: 3–5 big-number tiles (most critical KPIs with trend arrows)
2. Middle: 2–3 trend charts showing performance over time
3. Bottom: Detail tables or drill-down areas
4. Use consistent color: Green/Amber/Red for status, brand colors for categories

Step 5 — Data Dictionary

For each KPI, document:

• Definition: exactly how it's calculated
• Numerator / Denominator: if it's a ratio
• Inclusions / Exclusions: what counts and what doesn't
• Refresh frequency: real-time, hourly, daily, weekly
• Owner: who is accountable for this metric

Step 6 — Alert Rules

KPI	Condition	Severity	Action
[name]	Value < threshold for 2 consecutive periods	Warning	Notify owner via Slack/email
[name]	Value < critical threshold	Critical	Escalate to leadership

Inputs

Input	Required	Format
Business objectives	Yes	What decisions does this dashboard support?
Audience	Yes	Who will view this dashboard?
Available data sources	Yes	List of systems/databases
Existing metrics	Recommended	Current KPIs if any
Refresh requirement	Recommended	Real-time / daily / weekly

Output

## Dashboard Design — [Dashboard Name]

### Audience & Purpose
**Primary user:** Regional Sales Managers
**Decision supported:** Territory resource allocation
**Refresh:** Daily at 6 AM

### KPIs

| KPI | Type | Target | Green | Amber | Red | Owner |
|-----|------|--------|-------|-------|-----|-------|
| Monthly Revenue | Lagging | $1.2M | ≥100% | 85–99% | <85% | VP Sales |
| Pipeline Coverage | Leading | 3.0× | ≥3× | 2–3× | <2× | Sales Ops |
| Win Rate | Lagging | 28% | ≥28% | 22–27% | <22% | VP Sales |
| Activities/Rep/Week | Leading | 50 | ≥50 | 35–49 | <35 | Team Leads |

### Layout Wireframe
Row 1: [Revenue] [Pipeline] [Win Rate] [Activity Score]
Row 2: [Revenue Trend — 12 months] [Pipeline by Stage — stacked bar]
Row 3: [Rep Performance Table — sortable] [Territory Map — if applicable]

### Data Dictionary
[One entry per KPI with formula, source, refresh]

### Alert Rules
[Threshold-based notifications per KPI]

Definition of Done

• Max 7 KPIs per view
• Every KPI has a target, threshold, and owner
• Leading and lagging indicators balanced
• Visualization type justified for each data pattern
• Data dictionary complete
• Alert rules defined

Quality Criteria

• Data sources and assumptions are explicitly stated
• Calculations are reproducible from provided inputs
• Visualizations or tables have clear labels, units, and time ranges
• Caveats and confidence levels are documented for estimates

Verification (4C)

Check	Question
Correctness	Are formulas, aggregations, and statistical methods applied correctly?
Completeness	Does the analysis cover all requested metrics and time ranges?
Context-fit	Are the chosen metrics relevant to the business question being answered?
Consequence	If this data were used for a decision today, what blind spots remain?

Edge Cases

• Missing or incomplete data — Document gaps and their potential impact on conclusions. Provide ranges instead of point estimates.
• Outliers skewing results — Report with and without outliers. Document the decision to include or exclude.
• Changing data definitions mid-period — Split analysis at the change boundary and note the schema difference.

Changelog

• v1.0.0 — Initial release