Back to Skills

metrics-dashboard

Define and design a product metrics dashboard with key metrics, data sources, visualization types, and alert thresholds. Use when creating a metrics dashboard, defining KPIs, setting up product analytics, or building a data monitoring plan.

23,486stars2,489forksUpdated 7/11/2026

Security Assessment

Safe(100/100)
Security Score100/100

About metrics-dashboard

Designing a product metrics dashboard - the right metrics, data sources, visualization types, and alert thresholds - is what this skill guides, drawing on product-analytics frameworks and Ben Yoskovitz's Lean Analytics. It is meant for creating a dashboard, defining KPIs, setting up product analytics, or building a data monitoring plan, and it reads any supplied dashboards, analytics data, OKRs, or strategy docs first.

It grounds the work in domain concepts. Metrics are all measurable things, KPIs are the few key quantitative metrics tracked over a longer period, and the North Star Metric is a single customer-centric KPI that leads business success. Four criteria mark a good metric: understandable, comparative over time, expressed as a ratio or rate, and behavior-changing - the Golden Rule being that a metric that will not change how you behave is a bad metric. It also distinguishes eight metric types across vanity versus actionable, qualitative versus quantitative, exploratory versus reporting, and lagging versus leading, and lists five action steps for auditing and classifying metrics.

The instructions build the dashboard in layers: a North Star Metric, three to five input metrics that drive it, health metrics as guardrails, and business metrics for revenue, cost, and unit economics. Each metric is defined in a table capturing its exact calculation, data source, visualization type, target, and alert threshold. A layout sketch places the North Star at the top, input metrics with sparklines beneath, then health and business rows. A review cadence spans daily operational health, weekly input metrics, monthly North Star and business review, and quarterly recalibration, and alerting defines thresholds, recipients, channels, and expected response time. Tool recommendations are matched to context - Amplitude, Mixpanel, or PostHog for product analytics; Looker, Metabase, or Mode for SQL dashboards; Datadog or Grafana for operational health - and the final specification is saved as a markdown document.

FAQ

What metric layers does the dashboard use?

A North Star Metric, three to five input metrics that drive it, health metrics as guardrails, and business metrics for revenue, cost, and unit economics.

What makes a good metric?

Four criteria - understandable, comparative over time, a ratio or rate, and behavior-changing - summarized by the Golden Rule that a metric that will not change your behavior is a bad one.

What is defined for each metric?

Its exact definition or calculation, data source, visualization type, target value, and alert threshold, laid out in a table.

How often should metrics be reviewed?

Daily for operational health, weekly for input metrics and engagement, monthly for the North Star and business metrics, and quarterly for strategic recalibration.

Which tools does it recommend?

Amplitude, Mixpanel, or PostHog for product analytics; Looker, Metabase, or Mode for SQL-based dashboards; and Datadog or Grafana for operational health.

Install metrics-dashboard

Download and extract the skill files to your .claude/skills/ directory.

Quick Setup:

  1. Copy the skill folder to .claude/skills/
  2. Claude will automatically detect and use the skill