North Star Metric: How a Startup Picks the One Number to Steer By
A north star metric is the one number that best measures the value a company's customers get from its product, counted often enough to act on. It should move before revenue does, so it warns early, and every team should be able to influence it. A startup finds it at the moment a customer first gets real value, then splits it into inputs that each team owns.
Before a concert, the oboe plays a single A and the whole orchestra tunes to it. Skip that minute and each section can still be in tune with itself. The violins agree with each other, and so do the brass. The first chord still sounds wrong.
A startup's teams drift the same way. Marketing reports leads, sales reports pipeline, and each number looks healthy on its own slide. When two teams want the same budget or the same engineer, nobody has a shared number to settle it.
A north star metric is that tuning note: one number every team checks its work against. Two questions decide whether you've picked a good one. Does it count value your customers actually receive, and can every team see how its own work moves it?
North star metric tests
The oboe gets the job because everyone can hear it and nobody argues about the pitch. A candidate north star has to clear four tests before it earns the same trust:
- Customer value: it counts something the customer receives, such as reports delivered or orders shipped, rather than what the company collects.
- Early warning: it moves before revenue does, so a drop shows up while there's still time to fix it.
- Shared reach: every team can move it, even if some move it only indirectly.
- Memory: anyone in the company can say it, and how it's counted, without looking it up.
Revenue fails the first two tests. It records value after the customer has already decided to pay, and by the time it falls, the cause is a quarter old. Keep revenue on the dashboard. Just don't steer by it.
Examples by business type
Your business model decides the number, because it decides where your customer gets value. Treat these as hypothetical starting points to adapt:
| Business | North star | Why it counts |
|---|---|---|
| B2B software | Weekly accounts completing the core task | Use comes before renewal |
| Marketplace | Transactions completed per week | Both sides got what they came for |
| Service firm | Client deliverables accepted per month | Accepted work is value received |
| Online store | Repeat purchases per customer | A second order means the first one worked |
Each row counts a moment when the customer got what they paid for. None of them counts sign-ups or followers, because a number that measures attention can climb while customers quietly leave. Vanity metrics covers how to spot those.
Choosing a north star
A table gives you a starting point. Your own customers give you the answer, in five steps:
- Name the moment a customer first gets real value from you, in their words.
- Turn that moment into a count, over a window short enough to act on, usually a week.
- Check it against your history: the accounts that reached the moment often should be the ones that renewed or bought again.
- Split it into two or three inputs, such as new accounts reaching the moment and existing accounts returning to it, and give each input one owner.
- Write the definition down: the formula, the time window, what's excluded, and which system it comes from.
Most candidates fail at step three. If the customers you kept and the ones you lost scored about the same, the number isn't measuring value yet.
The inputs in step four make the north star usable day to day. Nobody can move the top number directly on a Tuesday afternoon. Marketing can move new accounts, and customer success can move returning ones. The weekly scorecard is where each owner reports on theirs.
Revenue operations owns that definition, because the number is only useful if every report computes it the same way. What is RevOps? covers who owns that job at a startup.
When it's too early: a company with a handful of customers doesn't need a formal north star. Talk to each one and count the ones who'd be upset if you disappeared. The Sean Ellis test turns that question into a survey once you have enough users to ask.
Common mistakes
A well-chosen number still fails when the company handles it badly. Four mistakes come up most:
- Counting your own effort: emails sent and calls made measure how hard your team worked, and a customer gets nothing from either.
- Picking a total: all-time sign-ups or total customers only go up, so they never tell you something changed.
- Changing it often: a north star needs time to show a trend. Revisit it once a year, or after a real change in strategy.
- Running several: two north stars put you back where you started, with each team tuned to its own note.
The number earns its keep when two projects want the same engineer. Ask which one moves the north star more, and by roughly how much. The answer won't make the decision alone, but it ends the debate about whose number matters. Put it at the top of your executive dashboard, with every other number below it.
Picking your north star
An orchestra tunes to the oboe because every player can hear it. Pick your north star the same way: a count of value your customers received, one that every team can hear move. If your teams can't agree on what a good week looks like, settle that before you build another report.
Want a second opinion on the number? My free diagnostic ends with a written roadmap either way, and you can run it yourself.
Tune to one note before you play louder.
Common questions
What's the difference between a north star metric and a KPI?
The difference between a north star metric and a KPI is scope. A KPI tracks progress toward one objective and usually belongs to one team. The north star metric is the single number those KPIs feed, so a company tracks many KPIs and steers by one north star.
Who should own the north star metric?
The north star metric belongs to the founder or CEO, because no single team controls it. Each input underneath it gets its own named owner, such as a marketing lead for new accounts, and the founder answers for the total.
Is a north star metric the same as an OKR?
A north star metric and an OKR do different jobs. The north star is a standing measure of customer value that holds for a year or more. An OKR sets a goal for one quarter, and a useful set of quarterly OKRs names which input to the north star the company is trying to move.
Does a B2B company with a long sales cycle need a north star metric?
A B2B company with a long sales cycle still benefits from a north star metric, as long as it counts customer value rather than deals. Accounts using the product every week, or deliverables a client accepted, move months before a renewal does and give the team time to act.