← ClaudeAtlas

metric-definitionlisted

Writes precise metric specs with grain, owner, source, formula, guardrails, and known caveats. Use when the user mentions metric definition, metric spec, KPI definition, "what is a session", "define X", North Star metric, or needs to disambiguate a metric across teams.
vermapragya/analytics-skill · ★ 0 · AI & Automation · score 72
Install: claude install-skill vermapragya/analytics-skill
# Metric Definition ## When to use this skill Use whenever a metric is being created, re-defined, or disputed. Triggers: - "Define our North Star metric" - "What counts as an active user?" - "Write a spec for revenue per user" - "Why is finance's revenue number different from ours?" - "Document this metric" Most metric arguments are actually definition arguments. Force the question into a spec. ## Required inputs | Input | Why it matters | |---|---| | Metric name | What stakeholders call it | | Business question | What decision the metric supports | | Owner | Who is accountable for accuracy | | Source tables | Where the underlying data lives | | Grain | Per user / session / day / order | ## Workflow 1. **Force the business question.** "Why does this metric exist?" If the user can't answer, the metric shouldn't exist yet. 2. **Pin down the grain.** "Active users per *what*?" — per day, per week, per month. The same metric name with different grains is three different metrics. 3. **Define the SQL.** Even if the user doesn't ask for it, write the canonical query. Vague definitions become inconsistent dashboards. 4. **Identify edge cases up front.** For each metric, ask: - What about deleted/banned users? - Internal/test users? - Refunded transactions? - Multiple devices per user? - Timezone (event time vs reporting time)? 5. **Define guardrails.** A primary metric without guardrails will get gamed. List 1-3 metrics that must not regress when this one m