north star metric

Also called: north star, NSM, north star metric framework, north star KPI

The one metric that best captures the value users get from the product and predicts long-term success. Teams use it to align decisions, not as the only number they watch.

A north star metric is the single number a product team agrees to move, chosen because it measures value delivered to users and tends to lead revenue rather than lag it. It gives every team the same answer to "is this change a win?", so a growth team, an onboarding team and a payments team stop optimizing numbers that pull in different directions.

A good north star usually passes five checks:

It reflects user value, not company value: people must be getting what they came for when it rises.
It leads revenue. When it grows, revenue and retention follow a few weeks or months later.
Teams can move it through their own work, via input metrics in a metric tree.
It is hard to game and can go down: a cumulative total ("all expenses ever added") only rises and tells you little.
It fits the product's rhythm. A weekly product needs a weekly count, not a daily one.

Most north stars combine breadth (how many users or accounts), depth (how much value each gets) and frequency (how often), for example "weekly active groups" or "orders delivered per week". Revenue itself is rarely a good north star: it is the result of value, arrives late and can rise while users get less (a price increase, a dark pattern at checkout).

The north star doesn't replace other metrics. It sits at the top of a metric tree of inputs that teams own, and next to guardrails (churn, support tickets, refund rate) that stop anyone from pushing it the wrong way.

Example

Halves (a bill-splitting app for flatmates, couples and trips) compares three candidates for its north star:

Candidate	Problem
App installs	Measures marketing reach, not value; a store feature can double it overnight.
Monthly active users (MAU)	A single user adding one expense alone gets little value from a shared expense app.
Weekly active groups: groups where at least 2 members added or settled an expense themselves that week	Captures the core value: people sharing costs together. Excludes server-created recurring bills.

The team picks weekly active groups and checks it against June, when a store feature brought a wave of new users:

MAU: 40,000 in May → 48,000 in June, a change of (48,000 − 40,000) / 40,000 = +20%.
Weekly active groups (average of the month's weeks): 12,000 → 12,240, a change of (12,240 − 12,000) / 12,000 = +2%.

MAU says June was a great month; the north star says most new users never shared an expense with anyone. Maya's conclusion: the partner promo brought installs, not value, and the onboarding team's job is to turn solo sign-ups into active groups.

Common mistakes

Picking revenue or installs. Both can grow while users get less value; the north star should measure the value that later produces revenue.
A cumulative total. "Total expenses ever added" can only go up, so it never warns you about anything.
Wrong time grain. A daily count for a weekly habit makes normal weekends look like crises.
No guardrails. Any single number can be pushed in harmful ways (spammy invites, forced notifications); pair it with churn and quality checks.
Changing it every quarter. A north star is useful because it stays put long enough for teams to learn what moves it.