Metrics Framework

Metrics Framework

Overview

A metrics framework connects what you build to what matters. Without one, teams optimize for whatever number is loudest. This skill covers AARRR (Pirate Metrics), HEART (Google), and custom metric trees — choose the right framework and build a measurement system that drives decisions.

When to Use

• Defining success for a new product or feature
• Building or redesigning your analytics system
• Aligning the team around a shared measurement model
• Moving from vanity metrics to actionable metrics
• Don't use for: one-time analysis (use experiment-analysis skill), daily monitoring (use dashboard-design)

Instructions

1. Choose a Framework

AARRR (Pirate Metrics) — best for growth-stage products:

• Acquisition: How do users find you? (traffic sources, signups)
• Activation: Do users have a great first experience? (aha moment, time-to-value)
• Retention: Do users come back? (D1/D7/D30 retention, churn)
• Revenue: Do users pay? (MRR, ARPU, conversion rate)
• Referral: Do users tell others? (viral coefficient, NPS)

HEART (Google) — best for UX and engagement-focused products:

• Happiness: User satisfaction (NPS, CSAT, satisfaction surveys)
• Engagement: Frequency and depth of use (sessions, actions per session)
• Adoption: New users of a feature (onboarding completion, feature activation)
• Retention: Returning users (cohort retention, churn rate)
• Task Success: Can users accomplish goals? (success rate, time-on-task, errors)

Custom Metric Tree — best for complex or unique products:

• Start with a North Star metric
• Break down into 3-5 supporting drivers
• Each driver has leading and lagging indicators
• Each indicator has a target and owner

2. Define Metrics Hierarchy

North Star: One metric that captures the core value your product delivers. Input metrics: Actions users take that drive the North Star (controllable). Output metrics: Results of those actions (lagging). Counter-metrics: What should NOT degrade while you optimize.

3. Document Each Metric

For each metric, document:

• Definition: Exactly how it's calculated
• Target: What good looks like
• Baseline: Current value
• Frequency: How often it's reviewed
• Owner: Who's responsible
• Source: Where the data comes from

4. Create Metric Health Checks

• 🟢 Metric on track, no action needed
• 🟡 Metric off track, mitigation plan exists
• 🔴 Metric off track, needs attention

Sample Output

Metric Tree for a Task Management Tool:

North Star: Weekly Active Teams (teams that update their project status at least once/week)

| Driver | Metric | Current | Target | Health | |--------|--------|---------|--------|--------| | Activation | Team setup completion rate | 45% | 70% | 🟡 | | Engagement | Updates per team/week | 3.2 | 5.0 | 🟢 | | Retention | Team W+4 retention | 60% | 75% | 🔴 | | Revenue | Conversion to paid | 8% | 12% | 🟡 | | Referral | Invites sent per active team | 1.5 | 3.0 | 🟢 |

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: uthumany
• Source: uthumany/uthy-legacy-os
• License: MIT
• Homepage: https://uthumany.github.io/uthy-legacy-os/

Install and usage instructions live in the source repository linked above.