North Star Metric: How to choose the top metric and why most people choose the wrong one

Spotify North Star has audition time. Airbnb has the number of nights booked. WhatsApp has the number of messages sent. Each of them has a single number that says, "If that number goes up, we're going in the right direction.".

Most teams want that number as well. And most people choose the wrong one.

Choose Revenue and start optimizing short-term sales at the expense of retention. Choose a DAU and chase the number of visits, forgetting the value of each visit. You pick an NPS and turn it into a bureaucratic procedure that nobody reads.

North Star Metric is not just the most important metric. It is a metric that simultaneously reflects the value to the user and leads to long-term business growth. When these two conditions are not met together, the metric doesn’t work like North Star.

## What is North Star Metric really about

The term was coined in the growth community, and it was used there for the first few years. He's now made it to grocery teams everywhere - and in the process has lost a bit of meaning.

North Star Metric is the only metric that best reflects the value a product creates for users while predicting long-term commercial success.

The two keywords here are “user value” and “predict.” Not “measuring revenue right now,” but predicting future success.

This is why Revenue itself is a bad NSM: it measures the past (what has already happened), not the future. Businesses can grow in revenue today and collapse tomorrow if users don’t get value.

But “the number of nights booked through Airbnb” is a good NSM: it only grows when hosts and guests get real value. No value, no reservations.

## Three criteria of the North Star Metric

### Criterion 1: Reflects value to the user

The metric should change if and only if the user gets something useful from the product.

“Users opened the app” does not reflect value. You can open it by accident. “Users completed the first project” reflects the value. That means the product helped them do something.

Check: Try to explain this metric to the user. If he understands what it is about and why it is important to him, good. If the metric is incomprehensible without the internal context of the company, it is bad.

### Criterion 2: Predicting long-term growth

NSM growth today should be correlated with business growth in 6-12 months. This is called the leading indicator, as opposed to the lagging indicator (which records what has already happened).

Revenue - lagging indicator. He says it already happened. Activation rate or engagement – leading indicators. They say what will happen to revenue in the future.

### Criterion 3: Managed by a team

The NSM needs to change in response to what the team is doing. If the metric depends mainly on external factors (season, market, competitors), it is difficult to manage it.

“Market share” is a bad NSM: it depends on competitors as much as it depends on you. “Activations per week” is better: it’s something that is directly impacted by product and marketing.

## Why Revenue is a Bad North Star

The most common wrong choice of NSM is Revenue or Profit. Intuitively, it seems logical: business lives for money, so Revenue should be North Star.

The problem is that Revenue is the result, not the cause. It grows when everything else works right: the product creates value, the users come back, the funnel converts. When you do Revenue NSM, the team starts optimizing Revenue — often at the expense of creating long-term value.

The classic symptoms of Revenue as NSM are:

• The team adds aggressively upsell mechanics that annoy users
• Cut functionality from a free plan so that new users don’t see value
• Focus on attracting new users while ignoring retention

As a result, Revenue grows a quarter or two, then churn begins to catch up.

## Why DAUs and MAUs Are Bad NSMs

The second most common erroneous choice is daily or monthly active users.

Problem: Active user is a very broad definition. The person came in and looked and did nothing - is he active? Technically, yes.

DAUs/MAUs are good as a component of NSM, but don’t say by themselves whether a user is getting value. Products with high DAUs and terrible UX exist: monopolists, network-effect products, products where there is no alternative.

If you want to count activity, count “quality activity.” Not “the user opened the app,” but “the user performed at least one key action.”.

## The formula for a good NSM

A good NSM is usually: [Number] [Users/Acts] [Received/Activated] [Performing Value] [Period].

Examples of real NSMs:

• Slack: Number of messages sent in teams per week
• Spotify: Listening to music per month
• Duolingo: Number of users who completed at least one lesson today
• GitHub: Number of pull requests per month
• Zoom: Number of successful minutes per week
• Notion: Number of blocks created by active users per week

Each has a specific action that reflects real use and value. Not just “the user is,” but “the user did what the product exists for.”.

## How to Choose NSM for Your Product

### Step 1: Determine the “moment of value”

Ask the question: when exactly does the user get what he came for?

For Spotify, listening to music. For Airbnb, when it comes to housing. For Notion, when you organize your notes so that they actually work.

It is a “aha moment” or “moment of value.” The NSM should measure the frequency of this moment.

### Step 2: Find the action that reflects it

The moment of value must be measured through a specific action in the product. “The user felt value” is not measurable. The user added 5 tasks to the project.

Sometimes one action fails. Then you need a composite indicator or you need to look for the best proxy.

### Step 3: Check the correlation with retention

A good NSM should correlate with long-term retention. This can be verified by data: users with a high metric value in the first week - do they stay after 30, 60, 90 days?

If there is a correlation, the candidate is good. If not, we need to look further.

### Step 4: Make sure the team can influence her

Brainstorm: What can the team do to make the NSM grow? If there are no ideas, either the metric is too abstract or the product is in the wrong field.

### Step 5: Choose one and fix it

Not two. Not “one primary and one secondary.” One. It's psychologically difficult -- it seems that one metric doesn't cover everything. But the point of NSM is in the focus. If there's two tricks, it's not North Star anymore.

## Input Metrics: How to turn NSM into action

The NSM itself does not explain what to do. This requires input metrics, the metrics that power the NSM.

If NSM is the number of completed projects per week, input metrics:

• Number of users who have started at least one project
• Percentage of users who add at least one member to the project
• Completion rate of the first step in the project

Working with input metrics is a real tactic. The NSM says where, input metrics says how to get there.

## Different NSMs for different product stages

NSM is not static. At different stages of product development, the focus is different.

Early stage (0→1): NSM is often the number of users who have received at least one valuable result. It is important to know if the product works at all.

Growth stage (1→N): NSM is usually the key action frequency. It is important that users return and use the product regularly.

Maturity stage: NSM often shifts to depth of use - engagement, expansion revenue, number of features used.

An NSM audit once a year is normal. Changing the NSM every quarter is a sign that the team doesn’t understand what it’s measuring.

## Traps for working with the NSM

### The "Gaming the Metric" Trap

When the NSM is fixed and everyone knows the purpose, it is tempting to optimize the metric rather than the actual value.

Example: NSM is the number of sessions per day. The team starts adding push notifications every 2 hours. Sessions are growing. Users are annoyed. Retention is falling.

Symptom: NSM is growing and business is not growing. This means that the metric has become a loophole.

Solution: periodically check the correlation of NSM with business outcomes. If it disappears, the metric is broken or broken.

### The NSM Obsession Trap

When you have an NSM, you are tempted to look at it. But the product is the system. The NSM grows, and somewhere in the system something breaks — and after a quarter it catches up.

Solution: NSM is the main metric, but not the only one. Counter-metrics (such as NPS or churn rate) are used as fuses.

### The "NSM for Reporting" Trap

Sometimes an NSM is chosen not because it is the right one, but because it looks good in a presentation to investors. This is an indicator of the maturity of the company.

A true NSM is chosen by the team that works with it, not the PR department.

## NSM and the role of designer

A designer doesn’t usually choose an NSM, it’s a product or CEO task. But a designer needs to know the NSM and understand how his work affects it.

Specifically, it means:

**How does this screen affect NSM? If not, either the task is wrong, or you need to find a connection.

**This recycling of onboarding will increase the activation rate, which will lead to an increase in NSM, because ... This is the language that business and product understand.

** If the decision is made and the NSM has not changed, you need to understand. Either the link was wrong, or the solution didn't work, or NSM was the wrong metric.

## Practical checklist: how to choose NSM

• You know the “moment of value” of your product – the moment when the user gets what he came for
• This point can be measured by a specific action in a product
• Users with a high value of this metric in the first 7 days are retained better after 30 days
• The team can name 3-5 specific actions that will raise this metric
• Metrics are not optimized bypassing without creating real value
• The whole product can be described as a growth machine for that one metric

## Example: Selection of NSM for EdTech Product

Let’s say there is an online platform for learning foreign languages.

** Candidates:**

• MAU – Many users come in but don’t learn
• Number of lessons purchased – Revenue proxy, not value
• The number of users who completed at least one lesson per week is close
• The number of users who showed progress (tested ≥ 70%) in a week is better because it reflects real learning

Select ** Number of users who completed at least one test with a score of ≥ 70% per week.

Checks:

• Reflects value? The user has really learned something.
• Predicts retention? Most likely yes – learning success correlates with a desire to continue.
• Managed? Yes – the quality of lessons, reminders, course structure, UX test.

Input metrics:

• Percentage of users who started the first lesson after registration
• Completion rate of each lesson
• Time between lessons (streak mechanics)

## Outcome

North Star Metric is not just the most important metric. This is the point of view of what is the success of your product.

If the NSM is chosen correctly, it creates alignment in the team: everyone knows where they are going. Designer, developer, marketer, product – all have different tasks, but one goal.

If the NSM is wrong, the team moves vigorously, but not there.

Choosing an NSM takes time and honesty. Don't pick something that grows easily. Choose what grows when and only when users get real value.

## How North Star Metric is Changing Design Prioritization

Before the NSM, the team answered the question "what to do next?" through intuition, politics and the volume of voices. After NSM, through data.

Specifically for the designer, this means that each proposed task should have an answer to the question “how does this affect NSM?”. Not as an abstract ritual, but as an honest filter.

Task: Redesign the user profile. The NSM connection is not obvious. Do users who fill out the profile use the product more often? If so, there's a connection. If not, the task is not a priority right now.

Task: Simplify step 3 onboarding. Step 3 - before the aha moment. Simplification will directly increase the number of activations, which directly grows NSM. High priority.

This type of thinking transforms the role of the designer: from the performer of tasks to the participant of strategy.

## Counterintuitive NSM: Unexpected Examples

Some products opt for NSMs that seem odd at first. But it is this unconventionality that reflects a deep understanding of the product.

Netflix is a watch for content. Not the number of subscribers (this is revenue-proxy), not the number of open applications (engagement without depth). It is the browsing hours – because only active browsing reflects why the user pays.

LinkedIn is the number of proven professional skills. At some point it was the NSM. LinkedIn makes money from advertising and subscriptions, while NSM makes money from profile content. Because full profiles → better matching → more value for both sides of the market.

HubSpot is the number of companies using CRM. Not revenue, not number of customers. Active use because HubSpot knew that inactive customers fall off the first renewal.

A good NSM reflects the actual activation of value, not the metadata of that activation.

## How to distinguish NSM from OKR

It's a lot of confusion. NSM and OKR are different tools with different functions.

NSM is a compass. One metric that says "in this direction." It doesn't change block by block. It reflects the fundamental nature of the product.

OKR is navigation. Objectives and Key Results are specific quarterly goals with measurable results. Change every 3 months. They reflect current priorities.

Connection: NSM is what key OKRs work for. Our NSM is the number of teams sending 100+ messages per week. This quarter our OKR is to activate 500 new teams to this level in 7 days instead of the current 14.

NSM does not replace OKR. It defines the vector that the OKR specifies.

## When NSM is Not Enough: Composite Metrics

For complex products, one metric sometimes doesn’t capture the whole picture. Then they use composite metrics.

Example - marketplace:

• Demand-side quality: CR from search to order
• Quality on the supply side: % of orders executed with a rating ≥ 4.5
• Composite NSM: Number of successful transactions (defined as an order rated ≥ 4.5 on both sides)

Family of Metrics is a small set of metrics (3-5), which together reflect the health of the product. Spotify uses a combination of time-in-app and user satisfaction.

If more than one metric is needed, it must be interrelated and independently verifiable. A set of unrelated metrics is not Family of Metrics, it's just a dashboard.

## Red Flags: Your NSM is incorrectly selected

Some signs that the current NSM is not working:

Metrica is growing, business is not growing. If NSM is steadily improving and revenue or retention is stagnant, the metric is out of touch with business reality. The connection you assumed doesn’t exist.

The team doesn't believe in metrics. ** If people think “so what” on the NSM plan, and not “wow”, the metric does not motivate. A good NSM generates energy: “We can move it!”

**Metric is easy to nail. If a developer can write a script in an evening that raises the NSM without actually improving the product, that’s a bad NSM.

There is no single understanding in the team. The product explains NSM in one way, the marketing in another, the designer in the third. If there is no consensus on meaning, there is no sense in metrics.

Metrics requires complex calculations. NSM should be explained in one sentence. If you want a 2-page methodology, it’s not North Star, it’s an analytical framework.

## AI and North Star Metric: How to Select and Verify the Main Metric

Choosing an NSM is a strategic decision that is rarely made, but comes at a cost when mistaken. AI helps in two stages: selection and verification.

### Prompt: generate candidates for the NSM

I do [a product description—what does, for whom, a business model].

Offer 5 North Star Metric options for this product.

For each option:
- Formulate the metric in one sentence
- Explain why it reflects value to the user
- Explain why it predicts long-term business growth
- Indicate the main risk of this metric (how it can be “gamified”)

In the end, what you recommend and why.

### Prompt: test the candidate for strength

When an NSM candidate already exists, check for possible holes

Our North Star Metric candidate: [formulation]

Check her for criteria:
1. Does it reflect real value to the user – or just activity?
2. Does it correlate with long-term retention? How do you test that with data?
3. Can it be grown without improving the product?
4. Is it clear to the whole team without context?
5. Do we have the leverage to grow it?

Evaluate each paragraph and suggest wording improvements if necessary.

### Prompt: Finding Input Metrics

Our North Star Metric: [formulation]

Offer 5-7 Input Metrics that power this NSM.

For each Input Metric:
- How does it affect NSM (direct communication)
- How to measure it.
- What design solutions can improve it

Prioritize the potential impact on the NSM.

### Prompt: explain NSM to the team

Sometimes the hardest part is not picking the NSM, but explaining it to the team in a way that everyone thinks in the same direction.

Our NSM: [formulation]

Write an explanation for three different audiences:
1. For CEO (1 paragraph, language of money and risk)
2. For the development team (1 paragraph, task and influence language)
3. For a new employee (1 paragraph, language of meaning and why)

In each explanation, a specific example of how NSM grows from a particular team action.