| name | pm-metrics |
| description | Define and track product metrics — North Star metric, input/output metrics, KPIs with targets. Creates metrics.json for the Command Center. Use when user says "/pm-metrics", "define metrics", "what should we track", "set KPIs", or needs to establish a measurement framework. |
/pm-metrics — Metrics Framework
Define your North Star, input metrics, and KPI targets.
Workflow
Step 1: North Star Discovery
Ask:
- "What's the one metric that best captures the value your product delivers to users?"
- "How often should it move?" (daily/weekly/monthly)
Guide toward a metric that is:
- Leading (predicts future success, not lagging)
- User-centric (measures value delivered, not vanity)
- Actionable (your team can influence it)
Examples by product type:
- SaaS: Weekly Active Users, Tasks Completed, Time Saved
- Marketplace: Transactions/week, GMV
- Content: Engaged Reading Time, Subscriber Growth Rate
- Developer Tools: API Calls/day, Projects Created
Step 2: Input/Output Metrics Tree
Build a metrics tree:
North Star: Weekly Active Users
├── Input: New signups/week (acquisition)
├── Input: Activation rate (% complete onboarding)
├── Input: D7 retention (stickiness)
├── Output: Revenue/user (monetization)
└── Guardrail: Support tickets/user (quality)
For each metric:
- Current value (or best estimate)
- Target (30/60/90 day)
- Data source (how we measure it)
- Owner (who's responsible)
Step 3: Generate metrics.json
{
"north_star": {
"name": "Weekly Active Users",
"current": 1240,
"target": 2000,
"target_date": "2026-06-30",
"frequency": "weekly"
},
"kpis": [
{
"name": "DAU",
"current": 1240,
"target": 2000,
"trend": "up",
"category": "engagement",
"source": "analytics"
}
],
"input_metrics": [...],
"guardrails": [...],
"last_updated": "2026-04-11"
}
Step 4: Wire to Daily Brief
The brief engine reads metrics.json automatically:
- Section 2 (KPI Pulse) shows current vs target
- Flags metrics that haven't moved in 7 days
- Celebrates metrics that hit targets
Step 5: Tracking Cadence
Suggest a tracking schedule:
- Daily: Check North Star + inputs
- Weekly: Full metrics review in retro
- Monthly: Adjust targets based on trends
Offer to schedule automated metric collection if data sources have APIs.
Update Mode
/pm-metrics update DAU 1350
Quick update a single metric value. Timestamps automatically.