Product metrics transform subjective opinions about product performance into
objective evidence. The right metrics create a shared understanding of whether the
product is succeeding and where to invest next. Metrics should drive decisions,
not just decorate dashboards. Every metric tracked should answer a specific

## Key Points

- **North Star Metric**: Identify the single metric that best captures the core
- **Pirate Metrics (AARRR)**: Track Acquisition, Activation, Retention, Referral,
- **Cohort Analysis**: Group users by signup date or behavior and track their
- **Funnel Analysis**: Map the steps users take toward key outcomes and measure
- **Leading vs. Lagging Indicators**: Track leading indicators (engagement,
- **Segmented Metrics**: Break aggregate metrics down by user segment (plan type,
- Define success metrics before building features, not after. Metrics that are
- Limit the number of metrics tracked actively. Three to five key metrics per
- Pair every efficiency metric with a quality counter-metric. Faster support
- Set targets based on benchmarks, historical trends, and strategic goals rather
- Instrument comprehensively but report selectively. Capture granular event data
- Review metrics weekly with the team and monthly with stakeholders.

Product Metrics

Core Philosophy

Product metrics transform subjective opinions about product performance into objective evidence. The right metrics create a shared understanding of whether the product is succeeding and where to invest next. Metrics should drive decisions, not just decorate dashboards. Every metric tracked should answer a specific question that influences a specific action.

Key Techniques

• North Star Metric: Identify the single metric that best captures the core value your product delivers to customers. All other metrics should either lead to or result from this metric.
• Pirate Metrics (AARRR): Track Acquisition, Activation, Retention, Referral, and Revenue as a framework covering the full customer lifecycle.
• Cohort Analysis: Group users by signup date or behavior and track their metrics over time to distinguish improvements from mix effects.
• Funnel Analysis: Map the steps users take toward key outcomes and measure conversion rates at each step to identify drop-off points.
• Leading vs. Lagging Indicators: Track leading indicators (engagement, feature adoption) that predict lagging outcomes (retention, revenue).
• Segmented Metrics: Break aggregate metrics down by user segment (plan type, geography, use case) to reveal hidden patterns and opportunities.

Best Practices

• Define success metrics before building features, not after. Metrics that are chosen after launch are susceptible to cherry-picking.
• Limit the number of metrics tracked actively. Three to five key metrics per team are sufficient; more creates diffusion of focus.
• Pair every efficiency metric with a quality counter-metric. Faster support response time means nothing if resolution quality drops.
• Set targets based on benchmarks, historical trends, and strategic goals rather than arbitrary round numbers.
• Instrument comprehensively but report selectively. Capture granular event data but surface only actionable insights.
• Review metrics weekly with the team and monthly with stakeholders.

Common Patterns

• Input Metrics → Output Metrics: Track controllable inputs (features shipped, experiments run) that drive desired outputs (retention, revenue).
• Health Scorecard: A single-page view of key product health indicators updated weekly with trend arrows and color coding.
• Experiment-Driven Metrics: Use A/B tests to establish causal relationships between product changes and metric movements.
• Customer Health Score: Composite metric combining engagement, satisfaction, and usage patterns to predict churn risk.

Anti-Patterns

• Vanity metrics that look good but do not inform decisions (total signups, page views, app downloads without engagement context).
• Measuring everything and acting on nothing. Dashboards without decisions are decoration.
• Goodhart's Law — when a metric becomes a target, it ceases to be a good metric. People optimize for the measurement rather than the underlying goal.
• Comparing absolute numbers across differently sized cohorts without normalizing.
• Ignoring metric seasonality and attributing normal cyclical patterns to product changes.
• Celebrating metric improvements without understanding whether they are statistically significant.