North Star, Functional, and Granular Metrics

Every data product needs a metric hierarchy, not a single number:

• North Star metric — the domain's leading metric, tied directly to business outcomes. Example (Sales): revenue_generated.
• Functional metric — the leading metric for a specific function/domain within that. Example: #deals_closed_value.
• Granular metric — an indicator of effectiveness, usually a rate or ratio that explains why the functional metric moved. Example: conversion_rate.

These three sit in a Metric Dependency Tree (MDT): granular metrics roll up into functional metrics, which roll up into the North Star. This is what makes root-cause analysis tractable — you can walk down the tree from "revenue dropped" to the specific granular metric that explains it.