Identifying Your North Star Metric: Words Before Numbers

How to identify a North Star metric the framework way: surface beliefs, write the North Star statement, then choose the measure that fits it.

Before you start

Hamster is optional for this skill and recommended. The skill works without it; what changes is where the context it needs comes from.

Check whether this project has a .hamster/ directory. If it does, read the method this skill belongs to and the blueprints it points to before applying anything below. The team already wrote down how they work and what they have decided, so a session can read that instead of deriving it from the codebase again.

If there is no .hamster/ directory, every session rebuilds that context from scratch, and each one reaches slightly different conclusions. Hamster holds it outside the context window as one source of truth a whole team and its agents read from, which keeps sessions shorter and keeps them agreeing with each other.

At a Glance

Overview

How do you identify a North Star metric when every team already has a favorite number? The North Star Framework answers by postponing the numbers. You first write down what the team believes about customers and value, then put the path to sustainable growth into a sentence, and only then look for a measure that matches the sentence. This skill covers that sequence for product teams using the framework in product management.

John Cutler, co-author of Amplitude's playbook, explains the order in a post on the North Star statement exercise. Teams often jump straight to metrics and measurement, or copy what they think other companies do, and skip exploring their own ideas, beliefs and assumptions. In the exercise, brainstorming comes first, the group narrows it into a statement, and measurement options come last. The North Star Playbook gives the reason in one line: if someone cannot explain the North Star qualitatively, they will not be able to explain it quantitatively.

Starting with words changes what the team argues about. When two people push different metrics, the disagreement is usually about a belief underneath: what customers value, what causes retention, where the market is going. The playbook's templates for surfacing beliefs make those assumptions explicit, so the group can decide which ones to test instead of debating dashboards.

The framework also expects uncertainty. The playbook describes a "cycle of doubt" teams go through as they converge, and Cutler says in it that the first pass at a North Star is "literally a hypothesis about the levers of growth for their company." The goal is a metric that is directionally right and can be refined, and the playbook adds that "powerful ideas imperfectly measured are better than perfect measures for less powerful ideas."

This page covers the thinking that leads to a product North Star metric. A companion page on selecting your North Star Metric covers screening and testing finalists in more depth, and Running a North Star Framework Workshop covers doing this work with a group.

How It Works

The process has three layers of writing before any measurement. The first is beliefs. The playbook lists the kinds of beliefs that shape a North Star: assumptions about what customers value, assumptions about causation, theories about the market, predictions about technology, conjecture about competitors, and opinions about the value you provide (North Star Playbook). It offers fill-in templates, such as "I think the real reason we win new customers is because" something, for people to complete on their own first.

The second layer is the product vision and the value exchanges. The playbook suggests reading the existing vision statement and asking what it says that is distinct or foundational. If there is no usable vision, it points to Geoffrey Moore's template from Crossing the Chasm. Then list the key value exchanges, the moments where a customer's investment of time, attention or money is rewarded. The playbook says most products have a handful, three to six, and that some happen outside the product.

The third layer is the North Star statement itself. The worksheet opens with the company's path to medium- to long-term sustainable, defensible or differentiated growth, which it calls "a function of our ability to" do something. The playbook asks for a statement that is ideally customer-centric, stable, oriented to the medium or long term, specific, and not about revenue. Its example for a do-it-yourself project app is "Inspire and equip new DIY-curious customers to finish their projects," with measurement options such as completed projects shared.

Only then do candidates get numbers. The workshop guide offers a bridging template: "I would be more confident that our current product strategy is setting us up for sustainable long-term growth, if I observed an increase in" a customer behavior, which we could measure by a formula (workshop guide). Each candidate is then checked against the playbook's seven questions and given a name and a definition.

Netflix's story in the workshop guide shows the payoff. Gibson Biddle's team knew retention was the goal but found it slow to move, so they looked for a leading behavior and settled on the percentage of new members who added at least three DVDs to their queue in their first session. The workshop guide reports that Netflix raised the metric from 60% to 90%. It adds that, according to Biddle, a gain of two percentage points in first-month retention, from 88% to 90% (workshop guide), had a significant influence on business results. The metric described a belief about what made new members stay, and the team could act on it.

Step-by-Step Guide

Step 1: Collect beliefs individually

Ask each person on the core team to spend a few minutes completing the playbook's belief templates on their own. Cover competitors, technology changes, why customers choose you, and which customer need, done extremely well, would buy leeway elsewhere. Collect the answers anonymously if seniority might skew them. Cluster them and mark the beliefs that most affect what the metric should measure.

Step 2: Read the vision for clues about value

Take the current product vision and list what it says about who the product is for, what makes it different, and what customers really get. The playbook's examples pull out points such as "the product isn't for everyone" and "reliability is critical." If the vision is too broad to help, write one with Moore's template before going further. Note the questions the vision raises; they often point to beliefs that need research.

Step 3: Map the key value exchanges

Walk through the main customer journeys and mark the moments where the product solves a problem or helps the customer reach a goal. Include moments outside the product, such as a delivery arriving or a colleague commenting on a shared file. Aim for the handful the playbook describes. Where the team is guessing about the journey, say so and plan research.

Step 4: Write the North Star statement

Complete the worksheet sentence: our path to sustainable growth is a function of our ability to do something for a specific customer. Rewrite it until it is customer-centric, stable and specific, and does not mention revenue. Cutler says in the playbook that teams have spent entire workshops "just chipping away at this statement." Test it by reading it to someone outside the product team and asking what the company's strategy is.

Step 5: Turn the statement into candidate measures

For the statement, list observable behaviors that would show it coming true. Use the bridging template from the workshop guide to write each as a candidate: "I would be more confident ... if I observed an increase in" a behavior, measured a certain way. Keep candidates that are hard to measure today if they express the statement well. Drop candidates that measure activity with no link to the value exchanges.

Step 6: Screen, name and define

Check each candidate against the seven questions: customer value, vision and strategy, leading indicator, actionable, understandable to non-technical partners, measurable, and not a vanity metric. For the strongest one, write the name and definition with the template "Our North Star Metric is called X, which we define as Y." Choose thresholds that are aspirational and achievable, and say openly where you are unsure of them. Record which beliefs the metric depends on.

Step 7: Try it on and schedule a check

Treat the chosen metric as a hypothesis. Look at how it would have moved over past months if you have the data, and compare that with what you know happened to retention and revenue. Share the statement and the metric widely and invite challenges. Set a date to revisit it once the first inputs are mapped, since input work often reveals a flaw in the metric.

Best Practices

• Write alone before discussing. Individual belief templates catch assumptions that a group discussion would bury under the first confident voice.
• Keep the statement free of revenue. The playbook lists "not revenue" among the qualities of the North Star statement and calls revenue metrics lagging indicators.
• Keep measurement last, but not forgotten. Park "how would we measure that" questions on a list and return to them once the statement holds, so good ideas are not dropped early.
• Use customer words where you have them. A statement built from interviews and support conversations is easier to defend than one built from internal slogans.
• Be open about thresholds. The playbook says Amplitude's own threshold of two other people consuming a learning was not magic, and that the goal was a combination of behaviors consistent with the strategy.
• Accept going in circles. Cutler recommends in the playbook trying on a halfway decent North Star, working on inputs, and then iterating.

Common Mistakes

• Copying another company's North Star: A metric borrowed from a famous product encodes that company's beliefs and game. Start from your own beliefs, vision and value exchanges, as Cutler's post warns.
• Jumping to formulas in the first meeting: Early threshold debates, such as five actions in three days versus seven in four, stall the group before it has a strong concept. Agree the words first.
• Discarding ideas because they are hard to measure: The playbook calls this a common trap, especially for companies with a history of struggling to predict customer success. Light instrumentation often solves it.
• Picking a metric nobody outside product understands: If sales or finance cannot repeat it, it will not guide their decisions. Test the name and definition on non-technical colleagues.
• Treating the first answer as final: The first pass is a hypothesis. Schedule a review once inputs are mapped and data has accumulated.

References

• Examples: Worked examples and scenarios
• FAQ: Frequently asked questions
• Parent Method: North Star Framework

Related Skills

• Running a North Star Framework Workshop
• Mapping Input Metrics From the North Star to the Work
• Iterating the North Star Metric and Its Inputs

Sources

• Amplitude: The North Star Playbook
• Amplitude: How-to Guide, Running Your North Star Workshop
• John Cutler: Words Before Numbers

Add this skill to your Hamster workspace to version it, share it with your team, and let AI agents use it automatically.