North star metric: how to choose one, with examples by business model

A north star metric is the one number that best captures the value customers get from your product and gives early warning of where revenue is heading. Teams typically place three to five input metrics under it, and those are the numbers they move day to day. If every planning meeting turns into the same argument, our product strategy consulting helps you set the shared rule that ends it.

Company examples appear only where the company published them, each with its date. The business-model examples are our own illustrations, so test them on your own data.

What is a north star metric?

A north star is a single rate, count or ratio that represents your product strategy and the value customers get from it. It is a leading indicator: when it moves, you expect business results to follow. Amplitude's guide to good and bad north stars places it above the many operational key performance indicators (KPIs) your team already tracks.

Amplitude's North Star Playbook gives it three qualities. It represents the value users get, it sits within the influence of product and marketing teams, and it is a leading indicator of revenue. Spotify's Confidence glossary adds a useful test: a good one measures value delivered, not activity captured.

In product management, the north star metric does one plain job: it settles trade-offs. Spotify's glossary calls it a common reference point when teams disagree about priorities. It does not replace your plan. Knowing how to build a product roadmap still matters, because the roadmap lists the bets and the metric shows whether they paid off.

North star metric examples by business model

The right number depends on how customers get value, so each business model points to a different number. Seat-based software counts finished workflows, while a marketplace counts completed transactions.

Amplitude sorts digital products into three games: attention, transaction and productivity. Spotify's glossary offers completed transactions for a marketplace and active projects per team for a collaboration tool. The table below is our own illustration, built on that logic, so treat each row as a starting hypothesis. Each model counts a different completed action, and each needs its own counter-metric.

The AI row shows why the wording matters. Message count looks like engagement, but a confused user can send more messages than a satisfied one. Counting accepted tasks moves the measure closer to the value delivered, and the cost counter-metric stops the team from buying that gain with expensive model calls.

Company north stars you can verify

Spotify, Amplitude and PostHog have each written about their own north star, and only Amplitude and PostHog gave a definition with a threshold. We left out any famous example we could not trace to the company.

• Spotify. Some lists state the Spotify north star metric as fact, yet the company's own wording is softer. As of October 2026, its glossary says that for Spotify, the metric "might be time spent listening".
• Amplitude. In a 2019 post, Amplitude described moving from weekly querying users to weekly learning users. A weekly learning user is active and shares a learning that at least 2 other people consume within 7 days.
• PostHog. Its North Star guide, first published in 2021 and updated in 2022, names Discoveries: analyzing an insight or a session recording for 10 seconds or more. The company says a customer growth metric felt premature at its early stage.

The Amplitude and PostHog statements are dated, so either company may measure something else today. The useful part is the shape of both definitions. Each names an action and a threshold, and Amplitude's adds a time window, which leaves little room for argument about what counts.

How to choose a north star metric in five steps

Choose it by naming the value customers get, turning that value into a countable action and testing the result on your own data. Five steps cover the work, and the last one checks the choice against real behaviour.

1. Name your game. Decide whether customers mainly give you attention, complete transactions or get work done. Amplitude's playbook calls this usually the first point of contention in a workshop.
2. Find the value moment. Write down when a customer gets what they came for. Amplitude's playbook describes a delivery app that found customers most valued deliveries with no issues, not simply early ones.
3. Make it countable. Give it a name, a qualifying action, a threshold and a time window. Then try it on a colleague outside the technical team. If they cannot follow it, rewrite it.
4. Run the checklist. Amplitude's North Star checklist asks whether the metric expresses customer value, represents your strategy, leads business results, can be influenced, is plain and measurable, and is not a vanity metric.
5. Test it on your data. Compare your candidates against retention or contract data. PostHog ran user cohort tests before choosing, and Amplitude ran correlations against its contract data.

Early products may not have enough data for step five. Pick a provisional metric, say so openly and return to it once you have learned how to find product market fit. Spotify's glossary makes the same point: a startup that starts with completed onboarding will outgrow it.

Which metrics make a poor north star?

Registered users and revenue usually make poor north stars. The first can rise while customers get nothing, and the second reports the past instead of giving early warning.

Amplitude's checklist marks daily active users and registered users as poor choices because they say nothing about what customers value. It says the same of average revenue per user, which tells you what already happened. Spotify's glossary adds that anything which can be inflated without creating value will be. In our own reading of four common candidates, only a measure of completed value comes close to passing all three tests.

Revenue still belongs on the dashboard. It just answers a different question. The north star should tell you early whether customers are getting value, and revenue should later confirm that you read it correctly.

Input metrics and counter-metrics

Input metrics are the levers teams can move, and a counter-metric warns when a gain is hurting something else. The north star itself should sit one level out of reach, so no single team can push it directly.

Amplitude's playbook describes inputs as typically three to five factors that most directly affect the metric and that you can influence through the product. Its troubleshooting chapter offers a heuristic with four dimensions. Picture a tool whose north star is key workflows completed each week. Multiply the four dimensions and you get that number.

• Breadth. How many active teams use key workflows.
• Depth. How many key workflows each team relies on.
• Frequency. How many runs per week each workflow gets.
• Efficiency. What share of those runs finish.

Each input gives a team something it can own, and this also suits experiments. Spotify's glossary says a north star rarely works as a success metric in an experiment, because it moves too slowly. A team on Spotify's search experience tests search-specific metrics instead, such as time to first play.

A counter-metric is not an input, because it feeds nothing. It should hold steady over time while the north star climbs. Spotify calls the experiment version a guardrail metric, which defines what you are trying not to break. Its examples include crash rates, latency and customer support contact rates.

North star metric vs OKR

A north star is a standing measure of value, while OKRs (objectives and key results) are time-boxed goals. You keep the north star across quarters and set OKRs that move its inputs.

Amplitude's playbook says the framework is not a goal-setting system, though it can be a strong foundation for goals. Its chapter on putting the framework into action suggests framing OKRs as the impact your bets will have on one or more inputs. Google's re:Work guide describes the other side: ambitious objectives, measurable key results and grades shared across the company. So a north star measures and stays, while an OKR sets a target and ends.

When to change your north star

Change it when your strategy changes, or when it stops pointing toward revenue. Do not change it because one quarter went badly, since a metric that moves with every result can no longer settle an argument.

Amplitude's playbook on changing your North Star lists the usual triggers: your market has changed, your business is at a different stage, or you cannot control the metric. Its own move came when its strategy shifted toward collaboration, and it says companies often get too attached to the old number.

Spotify's glossary calls a change a significant organizational decision, because it resets team priorities and experiments. Treat it that way. Write down the old definition and the new one, explain why you moved, and update the inputs and counter-metric at the same time.

How we help teams choose their north star

We start with the argument your team keeps having, then trace it back to the customer value underneath. From there we pick the one bet for the next stage and set the signal that will prove it right or wrong. For a bigger reset, you can hire product strategy consultants for a fixed-scope strategy sprint that ends in a written strategy your team can act on.

When the number has to hold up as the company grows, our startup scaling strategy work takes it further. The work is led by Roger Grekos, our co-founder for product, projects and scale. Every engagement is under an NDA, and the IP is always yours.

Frequently asked questions