metrics-review

A slash command for reviewing and analyzing product metrics, spotting trends, and turning raw numbers into a scorecard with recommended actions. It suits recurring weekly, monthly, or quarterly reviews, one-off investigations into a sudden spike or drop, and comparisons of performance against targets. When a product analytics tool is connected it pulls the key metrics plus comparison data and segment breakdowns; otherwise it asks the user to supply the values, comparisons, and context such as launches, incidents, or seasonality.

Metrics are organized with an explicit hierarchy. A North Star metric sits at the top, defined as value-aligned, leading, actionable, and understandable, with examples for collaboration tools, marketplaces, SaaS, content, and developer products. Below it are five to seven L1 health indicators mapped to the user lifecycle across acquisition, activation, engagement, retention, monetization, and satisfaction, and beneath those are L2 diagnostic metrics such as funnel step conversion, feature-level usage, and segment breakdowns. If the user has not defined a hierarchy, the skill helps identify the North Star and key L1 metrics first.

Trend analysis looks at each metric's current value, direction versus the previous period, comparison to target, rate of change, and anomalies, and it hunts for correlations and leading indicators. The generated review is structured into a short summary, a metric scorecard table, per-metric trend analysis with attribution, bright spots, areas of concern, recommended actions grouped as investigations, experiments, investments, and alerts, and a context and caveats section for data quality and comparability issues. It closes by offering follow-ups: deeper investigation of a metric, a dashboard spec for ongoing monitoring, drafted experiment proposals, and a reusable review template. Reference material defines common metrics like DAU, WAU, MAU, and the stickiness ratio.