# North star metric: what it is, 10 examples and a worked case

> What is a north star metric? North star metric examples for 10 product types, a test that exposes weak ones and a worked case on six months of data.

- Source: https://productovi.com/blog/north-star-metric-examples
- By: Sergey Bruh
- Published: 2026-09-30
- Topics: North star and metric tree

What is a north star metric? It's the single number a company steers by because it counts people (or teams, or groups) who **got the product's core value** in a period, at the rhythm they normally use it; the best north star metric examples all share that pattern. A marketplace counts completed bookings, not listings; a team tool counts teams that work together in it each week, not sign-ups; a bill-splitting app counts groups where at least two people share costs, not downloads. Below you'll find ten such examples by product type, a one-question test that exposes weak candidates, and a worked case on six months of data from Halves, a fictional expense-splitting app.

## What is a north star metric?

A [north star metric](/glossary/pa-north-star-metric) is the one number that best captures the value users get from the product and, because of that, predicts the business. It isn't the only number the team watches; it's the one the whole company agrees to steer by.

A good north star has five properties:

- **It reflects delivered value.** It rises only when users actually get what they came for.
- **It leads revenue.** People get value first and pay (or stop paying) later.
- **It breaks into inputs.** You can split it into smaller metrics that individual teams can move.
- **It's hard to buy.** One ad campaign shouldn't be able to jolt it in a day.
- **It has a written definition.** Two analysts computing it get the same number.

### The one-question test

For any candidate, ask: **can this number go up while users are worse off?** Downloads can double during a paid campaign in which nobody uses the product. App opens rise when a bug makes people check the same screen three times. If the answer is yes, the metric is a [vanity metric](/glossary/pa-vanity-metrics) or a volume metric: useful on a dashboard, but not as the company's goal.

A second check separates how much from how well: **would this number still move if the product didn't change at all**, say with a doubled ad budget or a holiday season? Sign-ups and MAU would. A north star should count units that actually got value.

## North star metric examples by product type

Generic examples; yours depends on where value appears in your product.

| Product type                   | North star example                                       | Tempting but weaker          |
| ------------------------------ | -------------------------------------------------------- | ---------------------------- |
| Team chat or collaboration app | Weekly active teams with at least 3 members posting      | Messages sent                |
| B2B project tool               | Weekly accounts where 2+ people complete tasks           | Paid seats                   |
| Two-sided marketplace          | Completed bookings per week                              | Listings, site visits        |
| Food delivery                  | Orders delivered on time per week                        | App installs                 |
| Online store                   | Customers placing a repeat order within 60 days          | Sessions, add-to-cart clicks |
| Subscription audio or video    | Subscribers with 3+ listening days a week                | Total minutes streamed       |
| Online learning                | Learners who finish at least one lesson a week           | Course sign-ups              |
| Fitness app                    | Members completing 2+ workouts a week                    | Daily opens                  |
| Budgeting app                  | Users who review or categorize spending at least monthly | Bank accounts connected      |
| Bill-splitting app             | Weekly groups where 2+ members add or settle an expense  | Downloads, MAU               |

Three things stand out in the table:

1. **The unit follows the value.** Where value needs more than one person (a team, a buyer and a seller, a group), the metric counts that unit, not individual users.
2. **The window follows the natural rhythm.** A chat app is judged daily or weekly, a budgeting app monthly. A daily window on a monthly product makes a healthy product look broken; the [DAU/MAU ratio](/blog/dau-mau-ratio) post shows this on real numbers.
3. **The "tempting" column is full of totals.** Minutes, messages and sessions are event totals: a few heavy users or an autoplay feature can inflate them while most people get nothing.

## A worked example: choosing the north star for Halves

Halves is an app for splitting shared expenses between flatmates, couples and friends on trips. The team is picking a company goal for the second half of the year from three candidates: MAU, new sign-ups and weekly active groups. First, ask what Halves's unit of value is. Nobody splits a bill alone; value appears only when a **group** of two or more people shares costs. So the candidate built from the value is:

**Weekly active groups:** groups where at least two members added an expense (server-created recurring bills don't count) or recorded a settle-up in the week.

Here are six months of data for all the candidates (staff accounts excluded):

| Metric                         | January | June   | Growth  |
| ------------------------------ | ------- | ------ | ------- |
| New sign-ups                   | 3,100   | 6,800  | +119.4% |
| MAU                            | 9,800   | 15,500 | +58.2%  |
| App opens                      | 61,000  | 90,200 | +47.9%  |
| Expenses added by people       | 41,200  | 47,300 | +14.8%  |
| Weekly active groups (average) | 1,480   | 1,655  | +11.8%  |
| Plus subscribers               | 410     | 498    | +21.5%  |

### Step 1: compute the growth rates

Growth is June divided by January, minus one:

- MAU: 15,500 ÷ 9,800 − 1 = 0.582, so **+58.2%**.
- Weekly active groups: 1,655 ÷ 1,480 − 1 = 0.118, so **+11.8%**.

In absolute terms, MAU rose by 5,700 users and weekly active groups by 175.

### Step 2: put value over reach

Divide the value metric by the reach metric to get weekly active groups per 1,000 MAU:

- January: 1,480 ÷ 9,800 × 1,000 = **151**
- June: 1,655 ÷ 15,500 × 1,000 = **107**

The ratio fell by about 29%. That alone doesn't mean Halves got worse: a ratio like this moves whenever the mix of users changes. A plausible first explanation is seasonal. Spring and early summer bring trip planners, and a trip group is busy for a week or two, then goes quiet. Its members count in MAU for the whole month but as an active group only during the trip, so more trip users means more MAU per active group even if every flat, couple and trip behaves as before. Only a split by group type and sign-up cohort can confirm it.

### Step 3: run each candidate through both tests

- **MAU and sign-ups** would rise with a bigger ad budget or a holiday season even if the product stood still. They're [volume metrics](/courses/product-analytics/volume-vs-quality-metrics/theory): for the dashboard, not for steering.
- **App opens** also rise when people struggle and check their balance again and again.
- **Plus subscribers** lag behind value (people pay for receipt scanning once they share many costs), and steering by revenue invites paywalls in the way of sharing. It becomes a monthly guardrail.
- **Weekly active groups** count the unit Halves exists to serve, and rise only when people actually split costs together. A bigger budget can bring people in, but it can't make two of them share a bill.

The recommendation: weekly active groups as the north star, MAU and sign-ups as volume metrics next to it, Plus as a guardrail. One honest limit: monthly totals can't separate a seasonal mix shift from a change in how groups behave. Next, split weekly active groups by group type and sign-up cohort, as in [how to calculate retention rate](/blog/how-to-calculate-retention-rate).

## The north star framework: from one metric to a metric tree

Nobody can walk into the office on Monday and "move weekly active groups". A [metric tree](/glossary/pa-metric-tree) splits the top into input metrics that one team can own. This pairing of one value metric with the inputs underneath it is what people usually mean by the north star framework. For Halves, this week's active groups come from three places:

- **Retained groups:** active last week and this week (owners: core expense features such as recurring bills).
- **Newly active groups:** new groups × share that get a second member × share that add a shared expense in week one (owners: acquisition marketing, onboarding and invites; see [activation rate](/blog/activation-rate)).
- **Returning groups:** active before, quiet last week, active again (owners: reminders, trip groups reused for a new trip).

Say last week had 1,600 active groups, and this week 1,360 of them were active again, 190 groups were active for the first time and 70 came back after a break:

1,360 + 190 + 70 = **1,620**, so +20 on the week.

A quiet week? Not quite: 1,600 − 1,360 = 240 groups went silent, and 260 new or returning ones replaced them. A flat top can hide a leaky bucket, and the tree shows which branch to open next. The full breakdown is in the lesson [The north star and the metric tree](/courses/product-analytics/north-star/theory).

![Halves's 1,620 weekly active groups break down into 1,360 retained, 190 newly active and 70 returning groups, each owned by a different team; 240 groups went silent behind the flat total](/blog/north-star-metric-examples/diagram.png)

## Mistakes that make a north star useless

### Picking the biggest number

MAU and downloads are large and usually go up. Size isn't a property of a good north star; sensitivity to value is.

### Counting events instead of units

"Expenses added per week" sounds close to value, but one busy trip or one heavy user can move it while most groups sit idle, and server-created bills inflate it without anyone opening the app. Count the unit that gets value (the group), with a qualifying action.

### Choosing revenue

Revenue is a result of value and it lags: a group that stops sharing costs today cancels its subscription next month. Keep revenue next to the north star and check it monthly. If the north star grows and revenue doesn't, look at pricing and packaging, not at value.

### Leaving it undefined

"Active groups" without a qualifying action, exclusions, a window and a time zone gives a different number to everyone who computes it. Write the definition under the chart; before launch it lives in the [tracking plan](/blog/tracking-plan-template).

## Key takeaways

- Good north star metric examples count units that got the core value, at the product's natural rhythm.
- Test every candidate: can it rise while users are worse off, or with no change to the product at all? If yes, it's a vanity or volume metric.
- The unit follows the value: teams, deals or groups when value needs more than one person.
- Compare candidates over the same months and normalize: at Halves, groups per 1,000 MAU went from 151 to 107 while MAU grew 58.2%; check the mix before blaming the product.
- Break the north star into a metric tree of inputs with owners; keep volume metrics and revenue beside it as context and guardrails.

## FAQ

### Can a company have more than one north star metric?

It can have one per product line when the lines do clearly different jobs, but inside one product, two north stars usually mean no north star. Keep one, and put the others in the tree or among the guardrails.

### Is revenue ever a good north star?

Rarely. Revenue lags value and is easy to raise in ways that hurt users, such as aggressive paywalls. It works best as a guardrail checked monthly next to a value metric.

### How often should the north star be reviewed?

Track it weekly or at the product's natural rhythm, and revisit the choice itself about once a year, or when the product's core value changes, for example after a move to a new kind of customer.

### What's the difference between a north star and an OKR?

The north star is a long-lived metric that describes value. An OKR is a time-boxed goal; a good quarterly key result is often a move in one of the north star's input metrics.

## Learn it hands-on

Start with the free lesson [The north star and the metric tree](/courses/product-analytics/north-star/theory), then pick and defend a north star yourself in the case [A north star for Halves](/courses/product-analytics/choosing-a-north-star/theory). For the bigger picture, read [what product analytics is](/blog/what-is-product-analytics).

## Related lessons

- [The north star metric and its KPI tree](https://productovi.com/courses/product-analytics/north-star.md)
- [Case: a north star for Halves](https://productovi.com/courses/product-analytics/choosing-a-north-star.md)
- [Volume, quality and Simpson's paradox](https://productovi.com/courses/product-analytics/volume-vs-quality-metrics.md)