The single metric a team chooses to prioritize, because it best captures the real value delivered to users.
In practice
The best-known examples: Airbnb tracked "nights booked" for years rather than sign-ups, Facebook tracked monthly active users rather than accounts created. The common thread: in both cases the metric captures value actually delivered, not an easily-inflated vanity action. A good North Star Metric has to answer one simple question: if it goes up and nothing else changes, is the business actually better off? If the answer isn't a clear yes, it's the wrong metric.
The formula
North Star Metric = users who get value × how often they get it × how much value each time — one number, three input levers
- Value delivered, not activity
The metric counts a moment where the user got what they came for — a night booked, a message read by a teammate, a report sent — never a proxy that can be inflated without anyone being better off (page views, sign-ups, sessions).
- Input metrics
The three factors are what teams actually work on: breadth (more users reaching the moment — activation), frequency (retention), depth (engagement or monetisation). The North Star is the output; nobody moves it directly.
- Leading, not lagging
Revenue is a result of value delivered months earlier; the North Star should move before revenue does, so that it predicts it. If your North Star is revenue, you're reading the rear-view mirror.
The test of a good North Star is a single question: if this number doubles and nothing else changes, is the business clearly better off? Sign-ups fail it (they could all churn). Nights booked passes it. So does "weekly active teams" for a collaboration tool — as long as "active" means doing the valuable thing.
Worked example
Choosing one for a B2B reporting tool
- Candidates: sign-ups per week (fails the test — they could all churn), reports created (closer, but a draft nobody reads is not value), monthly revenue (lags by months and says nothing about why).
- Chosen: reports sent to at least one recipient per week. It is the aha moment repeated; it requires a retained user; a team that sends more is a team getting more value.
- Decomposed: 1,200 sending accounts × 1.8 reports a week × 3.1 recipients each. Three teams, three inputs: onboarding owns the first, product owns the second, sharing owns the third.
- Two quarters later, revenue grows 30% — and the North Star had shown it a quarter earlier, when sending accounts crossed 1,500 while revenue was still flat.
A North Star doesn't replace the AARRR stages; it names which moment in them the whole company is optimising for. The stages then tell you where that moment is being lost.
Orders of magnitude
- The canonical examples: Airbnb's nights booked, Facebook's monthly (then daily) active users, Spotify's time spent listening, Slack's messages sent within teams, WhatsApp's messages sent. Each counts value received, and each was picked over a vanity alternative the company could have reported instead.
- One is the right number of North Stars. Two "North Stars" means the company hasn't decided; five means it has a dashboard. Teams keep their own input metrics underneath — that is the point of the decomposition.
- It changes as the business does: a marketplace's North Star at launch (listings created) is not its North Star at scale (transactions completed). Revisit it when the constraint moves, not every quarter.
How to improve it
- Start from the aha moment and ask what its repetition looks like: the North Star is usually that moment, counted across all users over a period.
- Run the doubling test on every candidate — "if this doubles and nothing else moves, are we better off?" — and discard the ones that fail, however convenient to report.
- Decompose it into three or four input metrics and give each one an owner. A North Star nobody can move is a slogan.
- Put it at the top of the one page the team looks at weekly, with the AARRR stage metrics underneath. The star says where you're going; the stages say what's in the way.
In the Tour
Do you know what percentage of users reach that moment?
- Yes, tracked precisely20 pts
- A rough sense of it7 pts
- No visibility at all0 pts
The Tour doesn't ask what your North Star is — but two of its questions are the ones a North Star is built from: "Do you know what percentage of users reach that moment?" (Activation: 20 for tracked precisely, 7 for a rough sense, 0 for no visibility) and "Do you track a retention rate?" (Retention). A team that answers 20 to both has the breadth and frequency terms of the formula above; the North Star is one multiplication away. A team that answers 0 to both cannot have one, whatever its slide says.
Questions people ask
Is revenue a good North Star Metric?
Almost never, for two reasons: it lags value by months, so it can't guide this quarter's work, and it can rise while users get less value (price increases, aggressive upsells) — right up until it collapses. Revenue is the goal; the North Star is the value creation that produces it. Businesses whose value moment is the payment itself (a marketplace's transaction) are the exception, and even they count transactions, not euros.
North Star Metric vs. OKRs — how do they fit together?
The North Star is the long-lived measure of value the company optimises; OKRs are quarterly commitments to move specific input metrics beneath it. A good OKR names an input (activation rate from 23% to 30%) and can explain, in one sentence, how it moves the star. OKRs that can't are activity dressed as outcomes.
Can a small startup have a North Star, or is it a scale-up thing?
It's more useful early, when the temptation to chase every number is strongest. A three-person team with one metric that means value — and a habit of asking whether each week's work moved it — makes better decisions than one with a full dashboard. It just needs an aha moment defined first, which is also the first Activation question of the Tour.