Executive Command Center Design

TL;DR

• Bottom line: An executive command center is a single-screen dashboard showing 8-12 KPIs with RAG status indicators, automated alert triggers, and action rules that turn metrics into decisions.
• Key tool/command: RAG threshold system — Green if metric > target, Amber if within range, Red if below critical
• Watch out for: Dashboard overload — more than 12 KPIs reduces actionability to zero.
• Works with: Any BI tool (Metabase, Grafana, Preset, Tableau, Looker) or custom-built with React + charting library.

Constraints

• Maximum 8-12 KPIs per dashboard screen. [src1]
• Every KPI must have a single accountable owner with an escalation path.
• RAG thresholds must use measurable data points, not subjective judgment. [src3]
• Limit automated alerts to critical events (red status only). Alerting on amber creates fatigue. [src5]
• Dashboard must load in under 3 seconds.
• Use symbols alongside colors for color-blind accessibility. [src5]

Quick Reference

Decision Tree

START: What type of command center?
├── Audience is CEO/founder only?
│   ├── YES → Single-screen overview with 5-8 KPIs covering all functions
│   └── NO ↓
├── Audience is executive team?
│   ├── YES → Overview + department drill-downs (one tab per function)
│   └── NO ↓
├── Audience is board of directors?
│   ├── YES → Quarterly snapshot with trend comparisons (simpler)
│   └── NO ↓
└── Company-wide?
    └── Simplified version with 3-5 company-level KPIs (no sensitive data)

KPI Selection:
├── Directly measures progress toward a company goal?
│   ├── NO → Remove it
│   └── YES ↓
├── Can an owner take action when it changes?
│   ├── NO → Move to drill-down page
│   └── YES ↓
└── Changes frequently enough for dashboard?
    ├── NO → Move to quarterly report
    └── YES → Include on command center

Step-by-Step Guide

1. Select KPIs Using the 3-Filter Method

Apply three filters: strategic alignment, actionability, and volatility. Choose 8-12 KPIs with a mix of leading and lagging indicators across revenue, acquisition, retention, product, efficiency, and team health. [src1] [src2]

Verify: 8-12 KPIs selected, each passes all 3 filters.

2. Define RAG Thresholds for Each KPI

Set measurable Green/Amber/Red boundaries for each KPI. Use percentages or ratios that scale with growth. Set amber as a buffer zone. Review thresholds quarterly. [src3] [src5]

Verify: Every KPI has Green/Amber/Red thresholds with measurable data points.

3. Design the Alert and Notification System

Create three alert tiers: Critical (red threshold, Slack DM + email, < 4 hour response), Warning (amber, dashboard badge only), Info (notable changes, weekly digest). Attach remediation checklists to critical alerts. Auto-escalate if unacknowledged after 48 hours. [src4] [src6]

Verify: Alert tiers defined, notification channels configured, remediation checklists attached.

4. Build Action Rules for Each Red KPI

Every red KPI must have a pre-defined action rule: IF metric crosses red THEN owner performs immediate diagnosis, implements corrective action within N days, and reports in next standup. Include escalation timeline. [src2]

Verify: Every red-threshold KPI has an action rule with owner, timeline, and escalation path.

5. Design the Dashboard Layout

Single screen with visual hierarchy: Row 1 financial KPIs (most important, top-left), Row 2 customer/product KPIs, Row 3 operational KPIs, Row 4 action items for red KPIs. Each card: large number + sparkline + RAG dot + owner initials. No pie charts. [src1] [src6]

Verify: Layout fits single screen, follows visual hierarchy, all KPIs visible without scrolling.

Code Examples

React: KPI Card Component

// Input:  KPI name, value, trend, threshold config
// Output: Styled card with RAG indicator and sparkline

function KPICard({ name, value, trend, thresholds, owner }) {
  const status = value >= thresholds.green ? 'green'
    : value >= thresholds.amber ? 'amber' : 'red';
  const icon = status === 'green' ? '✓'
    : status === 'amber' ? '⚠' : '✗';
  return (
    <div className={`kpi-card kpi-${status}`}>
      <span className="kpi-status">{icon}</span>
      <h3>{name}</h3>
      <p className="kpi-value">{value}</p>
      <Sparkline data={trend} />
      <span className="kpi-owner">{owner}</span>
    </div>
  );
}

Python: Alert Threshold Evaluator

# Input:  KPI data dict with current values and threshold config
# Output: List of alerts for KPIs that crossed red threshold

def evaluate_alerts(kpis, thresholds):
    alerts = []
    for kpi_name, current in kpis.items():
        config = thresholds[kpi_name]
        if config["direction"] == "higher_is_better":
            if current < config["red"]:
                alerts.append({"kpi": kpi_name, "value": current,
                    "threshold": config["red"], "severity": "critical"})
        else:
            if current > config["red"]:
                alerts.append({"kpi": kpi_name, "value": current,
                    "threshold": config["red"], "severity": "critical"})
    return alerts

Anti-Patterns

Wrong: Dashboard with 30+ KPIs

Showing every available metric because stakeholders each want their metric visible. Executives glance for 2 seconds, absorb nothing, and stop using it. [src1]

Correct: 8-12 KPIs with drill-down

Limit the command center to 8-12 KPIs that pass the 3-filter test. Everything else goes on department drill-down pages.

Wrong: Subjective RAG assignment

Letting people manually assign green/amber/red based on feeling. Everything is green until it is suddenly, catastrophically red.

Correct: Data-driven RAG thresholds

Define measurable thresholds before launch. RAG is calculated automatically from data. No human judgment in status assignment. [src3]

Wrong: Alerting on everything

Sending notifications for every amber and red change. Within a week, the channel is muted. A real critical alert is missed.

Correct: Alert only on red, digest for amber

Reserve push notifications for red-threshold crossings. Amber changes go in weekly digest. This preserves signal value. [src5]

Common Pitfalls

• Vanity metrics on the dashboard: Showing total users instead of active users. Fix: every KPI must answer "so what?" [src2]
• No data freshness indicator: Executives make decisions on stale data. Fix: show "last updated" per card, highlight red if > 24 hours old. [src4]
• Building before defining KPIs: Jumping to tool setup before deciding what to measure. Fix: complete KPI selection and thresholds on paper first.
• One-size-fits-all dashboard: Same view for CEO, board, and all-hands. Fix: create 2-3 views with different KPI subsets. [src6]
• No owner accountability: KPIs exist but nobody is responsible when they go red. Fix: display owner name on every card, auto-escalate unacknowledged red status. [src2]

When to Use / When Not to Use

Important Caveats

• RAG thresholds should be reviewed quarterly as the company grows
• Alert systems require ongoing tuning — start conservative and add gradually
• Dashboard adoption requires executive sponsorship — if the CEO does not use it, nobody will
• Consider time zones for alert routing in distributed teams