── ── Startups
North Star Metric
The North Star Metric (NSM) is the single metric that most directly measures value delivered to customers and predicts revenue over time. Popularized by Sean Ellis and Amplitude. Revenue is the goal; the NSM is the leading indicator that predicts it early enough to act — picking revenue itself produces a lagging dashboard, not a steering wheel.
Run North Star Metric on a real problem
Bring something you're actually deciding — free, in the browser.
How it works
1. Articulate customer value in customer units — the customer's outcome, not your product's mechanism. 2. Generate 3–5 NSM candidates — each proxies that value as a measurable metric. 3. Apply the 3 criteria (Amplitude §2): (a) Customer value? (b) Strategy fit? (c) Leads revenue? All three required; two-of-three = supporting metric only. 4. Time-shifted correlation — does the candidate lead revenue over 6–12 months? 5. Perverse-incentive stress test — could the team game this in a way that hurts customers? 6. Pick one. Put it on the wall. Explicitly name supporting metrics as supporting, not NSMs. 7. Re-evaluate quarterly — early stage → engagement; growth → retention; scale → revenue-adjacent.
When to use it
- ** teams are optimizing conflicting metrics
- dashboard has 30+ metrics with no priority
- a leading indicator of revenue is needed
- someone says "NSM," "OMTM," "what should we optimize," or "we measure too many things"
- an AI-native product is chasing sign-ups / prompts / demo plays and needs an activated-value metric that survives high inference/capex costs and AI-adoption churn
When not to use it
** product has no customers (pre-PMF → use lean-startup); single-team execution in a mature business; genuinely conflicting strategic objectives (the strategy needs work, not a metric).
Worked example
Choosing a North Star for an AI Product — Activated Value vs. Vanity Metrics…
A worked example applying the NSM Audit to the class of AI-native products (assistants, copilots, and agents) that proliferated after ChatGPT's late-2022 launch. The composite pattern below reflects widely-reported dynamics of the 2023–2026 AI product wave, not the internal dashboard of any one named company.
Install this skill (free, MIT)
npx skills add deciqAI/knowledge-skillsUseful? Star the repo — stars help other builders find it.
FAQ
What is a North Star Metric with examples?
The single metric that best measures the value customers actually receive and predicts revenue early enough to act on: nights booked for Airbnb, weekly active teams for Slack, orders delivered for a marketplace. It's chosen so that moving it necessarily means customers got more value.
Why not just use revenue as the North Star?
Revenue is the goal but a lagging indicator — by the time it moves, the causes are months old. A good NSM leads revenue: value delivered today becomes retention, expansion, and referral revenue later. Steering by revenue is driving by the rear-view mirror.
How do I choose the right North Star Metric?
Test candidates against three criteria: it measures delivered customer value (not activity), it predicts revenue, and every team can influence it. Then guard against Goodhart's Law — pair it with counter-metrics so optimizing the number can't drift away from the value it represents.
Related mental models
Organic growth builds value from within (product, operations, retention, unit economics).
International expansion forks into two models: Spirit Expansion (精神出海) — product/IP travels digitally or via exported goods, no local presence required — and Physical Expansion (实体出海)…
Two distinct questions decide early-stage growth: (1) PMF — do enough users love the product they'd be very disappointed without it?
Ramen profitability (Paul Graham) = the startup makes just enough to cover the founders' basic living costs.