North Star metric: how to choose yours

A North Star metric is the one number that best captures the value your customers get from your product. Choose one that reflects that value, leads to revenue over time, and that your team can move through its own work. Then point every growth experiment at it or at one of its inputs.

What a North Star metric is for

A startup can track dozens of numbers. Most of them matter a little. The North Star is the one that matters most, the number the whole team agrees to grow.

It gives you a shared answer to a hard question: is this week's work making the business better? When people disagree about priorities, the North Star breaks the tie. It also makes progress easy to explain, to the team and to anyone backing the company.

What makes a good one

A good North Star passes three tests.

  • It reflects value delivered to customers. It goes up when customers get what they came for, not just when they visit or sign up.
  • It leads revenue. If it keeps growing, revenue should follow. It moves before revenue does, so it gives you an earlier signal.
  • The team can move it. Product, marketing and sales work should all be able to push it, within weeks rather than years.

Watch out for vanity numbers. Total signups or page views can climb while customers get nothing. Revenue alone is also a weak choice, because it lags behind the value that creates it and tells you little about what to fix.

Examples by business type

The right metric depends on how your customers get value. Here are common patterns, described in general terms.

  • A marketplace: completed transactions between buyers and sellers. A listing or a visit is not value. A finished deal is, for both sides.
  • Subscription software: weekly active accounts that complete the core action, such as sending a report or closing a task. This shows customers are getting the job done, which keeps them paying.
  • A media business: time spent engaged with content by returning readers or viewers. Return visits show the content is worth coming back for, which drives both ads and subscriptions.

Notice that each one counts a real outcome, not an activity on the way to it. Two companies in the same category can still choose different metrics, because their customers get value in different ways. Start from the moment a customer would say the product worked for them, and count that.

How it connects to experiments

A North Star is too big to move directly. Break it into its inputs: the smaller numbers that, together, produce it. For subscription software, those might be new accounts, the share of new accounts that reach the core action, and how often active accounts come back.

Every experiment should move the North Star or one of those inputs. Write the metric into the hypothesis so the link is explicit. If you cannot say which input an experiment moves, it probably does not belong in the backlog.

This also helps with scoring. An experiment that moves a strong input of your North Star deserves a higher impact score than one that moves a side number.

Revisit it, but not often

Pick a North Star and keep it long enough to learn from it. Changing it every month leaves the team without a direction, and makes old results hard to compare with new ones. Revisit it when your business model changes, or when you find it rising while customers are not better off.

Keep reading

Rank all 19 channels for your company, free.