North Star MetricA customer-value atlas

One measure
of customer value.

Choose a North Star that describes what customers achieve, then surround it with a small set of measures your team can influence.

WorkshopKit / fictional supplied contextProductivity game
One proposed North StarWeekly workspaces completing a workshop with a recorded next action
Input 01

Publish

Weekly publishing workspaces
Distinct workspaces with workshop_published in the week.

Input 02

Complete

Published-workshop completion rate
Published weekly cohort completed within 14 days. Report only matured cohorts.

Input 03

Record an action

Next-action capture rate
Completed workshops with a same-workshop action_recorded within 24 hours.

Input 04

Repeat value

Repeat qualifying-workspace rate
Last week’s qualifying workspaces that also qualify this week.

A proposed influence map, not a proven causal model. The saved output defines one weekly count and four supporting measures; no real analytics were connected.
North Star Metric: visual rendering of the saved fictional demonstration output
Fictional demonstration. A visual rendering of the actual saved North Star Metric output from the source-guided Codex preparation run. Targets and proposals are not achieved business results. Select the image to open the full-resolution image.

For teams deciding what to measure

Useful for product leads, founders and operations teams defining a measurement framework. Supply the business model, customer value, vision, segments, current measures and available events.

Receive an Attention, Transaction or Productivity classification; one customer-centered North Star reviewed against seven criteria; and three to five supporting input metrics.

A proposal you can inspect

The output is reasoning and definitions. It does not connect to analytics, instrument a product or build a dashboard. Leading-indicator claims remain hypotheses until checked with real data.

Use through your assistant

Extract the ZIP and ask an AI assistant to read SKILL.md, or use its supported Skill installation method. This is an instruction-only Skill. No additional service API key or code runtime is required for supplied-input reasoning; model/subscription costs can apply.

Keep the evidence boundary clear

Metric labels do not create measurement data. Confirm event definitions, denominators, segmentation, time windows and potential gaming. Use only business information you may share.

The controlled Codex example passed six mechanical structure checks. No live analytics, causal relationship, business improvement or native-client installation was tested.