← ClaudeAtlas

north-star-input-treelisted

Build a North Star Metric and decompose it into 3-5 controllable input metrics that sum/multiply to it — the Amplitude framework as a standalone build. Covers NSM selection criteria, the input-metric tree, finding the weakest input to focus on, and the anti-patterns (revenue/vanity NSMs). Makes a North Star *actionable* instead of a scoreboard. Used by metrics-architect.
VandanaAjayDubey111/great-pm · ★ 3 · AI & Automation · score 74
Install: claude install-skill VandanaAjayDubey111/great-pm
> **Provenance.** great-pm-original, authored 2026-05-29, grounded in the > cited sources (Amplitude North Star Playbook, Open Practice Library, > Alex Morcillo). Host agent: metrics-architect. # North Star + Input-Metric Tree — making the one number actionable A North Star Metric (NSM) on its own is a **scoreboard**: it tells you the score but not which move to make. It becomes *actionable* only when decomposed into **3–5 input metrics** — the levers teams directly influence that *combine* (sum or multiply) to produce the North Star. Sean Ellis coined the NSM concept; **Amplitude formalized the input-metrics tree** that turns it from a vanity dashboard into a decision-making instrument. This skill builds both halves: select an NSM that survives the criteria, then decompose it into a tree the team can move. The defining property: a North Star expresses the **value customers get from the product**, framed as a *leading* indicator of business success — not the business success itself (that's revenue, a lagging output). The input tree is what lets a decentralized team find the *weakest input*, run experiments there, and check whether the North Star moved. ## 1. Selecting the North Star — the criteria A candidate NSM must satisfy *all* of these. Score each candidate against the checklist; the one that passes cleanly wins. | Criterion | What it means | Failure mode it catches | |-----------|---------------|--------------------------| | **Expresses customer value** | Captures