GUIDE 2024

PMs & Metrics: North Star Metrics

Because product managers fill in the white space between customers, the business, and the development team, product managers are always flooded with qualitative information – ranging from complaints to quotes to praise to anecdotes.

To successfully manage the chaos, product managers rely on metrics to aggregate information. That way, they can easily digest and diagnose the flood of information that they’re getting.

Yet even then, product managers can still be swamped with too many metrics. Given that product managers must prioritize, how do they make sense of the world around them?

The most successful product managers rely on a single north star metric (NSM) and a counter metric.

A north star metric measures the most critical outcome of success for the product.

A counter metric ensures that this improvement didn’t come at the expense of an equally important outcome.

Below, I’ll discuss how to select a north star metric that makes the most sense for you and your product.

Then, in a later article, I’ll discuss how to select a compelling counter metric, as well as how to pull them both together so that you can enable your stakeholders to strive towards true success.

North Star Metrics

Pioneers of the past used the North Star in the night sky to navigate their way through treacherous, uncharted territory.

Similarly, product managers are pioneers of uncharted territory, and they need to find their way to success by looking for clear signals in a sea full of noise.

That’s where the north star metric comes in.

The north star metric is the most critical, measurable, actionable business outcome for your product. In other words, it’s the one thing you can use to orient yourself towards product success – and therefore, success for your entire company.

Let’s break down each of the adjectives I mentioned.

The North Star Metric Must Be Critical

The north star metric should be critically important to your product’s mission, which then feeds into your company’s business model.

If your product is meant to make it easier for new users to onboard to the rest of your company’s product portfolio, it doesn’t make sense for you to anchor on the number of new users as your north star.

Rather, in this case, it makes far more sense for you to anchor on the total time it takes for new users to onboard, because that aligns with your product’s mission.

In other words, when selecting a north star metric, you need to be comfortable with acting as though it were the only piece of information you had about your product.

Based only on changes to your north star metric, how confident are you that you’re moving in the right direction? If you’re very confident, then you’ve picked a good north star metric. If you’re left with lots of questions and doubt, then you haven’t picked a good one yet.

Why am I purposefully limiting us to only a single metric for success? Couldn’t a product have many different levers to reach success?

It’s true that products have a variety of impacts. The problem is that if you use more than one metric, you’ll wind up getting distracted by noise.

Take it from my own personal experience! When I was an associate product manager for a SaaS B2B product, I used to have a dashboard that had 20 different metrics on it. I tracked metrics like conversion rates, new users, and churn rates.

Every day, I’d see some metrics go up and some metrics go down. And every day, I’d chase the metrics that had performed poorly. Rather than focusing on the one most important thing, I tackled a bunch of things that ultimately didn’t yield impact.

Because I hadn’t selected a single metric as the most important metric, I ran around chasing a bunch of self-inflicted “fires” that didn’t actually move my product towards the success of my customers and towards the success of my business.

I wasted engineering time, design time, product management time, and customer success time in going after random fluctuations rather than focusing on the sustainable growth of a single core metric.

The key is that while a product could have many different levers for success, these levers should all aggregate to the north star metric.

As an example, say that you own the onboarding product, and you’ve decided that your north star metric is “time it takes for a new user to successfully set up their user settings.”

You could measure a variety of things, for example:

  • The time it takes for a new user to move from the landing page to the user settings page.
  • The number of clicks they take within the user settings page.
  • The number of errors that appear when users attempt to save invalid settings.
  • The bounce rate for users on the settings page.
  • The percentage of users who reach the settings page that fails to customize the settings away from the default settings.

Each of these key performance indicators, or KPIs, might be something that you can control from optimizing your product.

But at the end of the day, all of these fold into the broader metric of “time it takes for a new user to successfully set up their user settings.”

So, now we understand why the north star metric needs to be the single most critical metric for your product.

As an aside, one thing to keep in mind is that the north star metric should not be a vanity metric. In other words, it shouldn’t be something that will grow over time no matter what you do.

For example, the total of signups or registrations that you have is not a north star metric, because it will always go upwards – it can’t possibly go down. People can’t “un-sign-up” from your product, so you don’t want to use that as your north star metric.

The North Star Metric Must Be Measurable

You’d be surprised at how many people make the mistake of selecting a north star metric that can’t be measured at all.

For example, let’s say that your product is a mobile app that focuses on increasing spoken English fluency for children. You decide that you want to measure an increase in the total of distinct words that each child can say without assistance.

How, exactly, would you measure this metric?

You can’t just turn on the microphone and record the child’s voice 24/7 without permission.

You can’t send a researcher to every single household that uses your product for a longitudinal spoken word study.

You can’t surface a sentence on the screen for the child to read out loud, because that would be assisting them with visual cues.

You can’t ask either the child or the parent to count out the number of words that the child knows how to speak because it’s highly likely that they’ll self-report incorrectly.

The problem doesn’t lie in the measurement of the metric, but rather in the selection of the metric itself. If you don’t have a good proxy for measurement, you can’t use that metric.

Let’s be absolutely clear – metrics are not reality. Metrics are human inventions that help us to approximate reality.

There are simply some things that you can’t measure because metrics are only tools to help us get close to reality.

So, let’s be less strict about the specific metric here, and loosen some of the requirements. We can use approximations to the ideal metric that we wanted to move.

Here are a couple of alternatives that can move us in the right direction:

  • Measure an increase in percentage of words that a child understands on-screen.
  • Measure an increase in the of distinct words that a child uses to respond to a prompt for a speech recording.

With the first alternative, we now have something that we can actually measure.

As an example, we could design a daily 10-question quiz on our app where the app says a pre-recorded word (e.g. “banana”), and we give the child a multiple choice of 4 words that they can choose from (e.g. “balloon”, “banana”, “pan”, “dog”).

We can measure whether the child is increasingly more successful in selecting the word that matches the recording. It’s not a perfect match with our unmeasurable ideal metric, but it gives us a directional sense that our app is doing what we want it to do.

With the second alternative, we also have something that we can actually measure.

As an example, we could design a weekly prompt on our app where the app asks a pre-recorded question, and asks the child to respond. For example, we could ask “what was your favorite moment today?” or “tell me about one of your friends.”

There, we could parse the child’s recorded response and count the number of distinct words that they used.

Again, it’s not perfectly aligned with the mission of our app, but it moves us in the right direction!

If you can’t measure the metric, you can’t determine whether you’re headed in the right direction.

As a reminder, when pioneers used the North Star in the past to navigate, they didn’t expect the North Star to be as exact as the GPS maps that we use today. Rather, they used it to move directionally towards their desired destination.

The North Star Metric Must Be Actionable

Finally, you need to be able to act on your north star metric. If you can’t take action on it, then it’s not providing you or your product team with value.

After all, if your data can’t inform your decision-making, then your data isn’t actually helping you succeed.

Here’s a real-life example: every B2B product would ideally use “revenue generated” as its north star. After all, revenue is the key goal, and revenue is measurable.

But, problematically, sales cycles in B2B can take a long time. It might take 6 months, 12 months, or even 18 months for a deal to close. How would you know whether your feature actually meaningfully moved revenue upwards if it takes you years to identify whether you made an impact?

Again, metrics are simply approximations for reality. Here, we can swap in our revenue metric with a leading indicator or something that we believe is strongly correlated with our ideal metric.

For example, say that you’re a product manager for Salesforce, which sells B2B products. If you’re in charge of some functionality for their CRM (customer relationship management software), you would ideally like to measure revenue.

But, you’ve noticed that the customers who typically renew their contracts are the ones who have the most active users on your product. Or, you might notice that customers with more records in your product are more likely to buy other features as well.

Now you have two potential leading indicators to use. You can use the number of active users, or you can use the of records. Both can be measured in real-time without needing to wait for a renewal contract to be signed. Say you decide to use “number of active users” as the north star metric for your product.

You now have the capability to run experiments in real-time to see whether you can move your north star metric upwards. You can launch changes in wording, changes in user flows, or even conduct A/B tests for a feature that you’re hoping to roll out broadly across the customer base.

And, you can see in real-time whether the of active users suddenly drops. Say, for example, that there’s an outage on the Salesforce platform, which means that no one has been active on the last day.

North star metrics are actionable because they trigger an immediate investigation – why don’t we have any active users? And how can we get that fixed as quickly as possible?

That’s one reason why user engagement metrics such as daily active users (DAU), weekly active users (WAU), and monthly active users (MAU) are commonly selected as north star metrics. User engagement metrics are leading indicators towards the ultimate goal of “increase revenue”, and these metrics can quickly be acted upon.

Closing Thoughts

Product management is a balance of art and science. While you can be highly scientific in measuring particular metrics, you still need to make subjective decisions on what a good north star metric should be.

A good north star metric helps you focus. Therefore, it needs to be a critical, measurable, and actionable metric that you can align your team on.

Many solid technology companies rely on north star metrics to guide themselves to success. Airbnb, Google, and Dropbox all rely on north star metrics.

In our next article, we’ll discuss how to use counter-metrics to ensure that you’re still talking of what matters and that you’re not taking detrimental shortcuts.

Have thoughts that you’d like to contribute around north star metrics? Chat with other product managers around the world in our PMHQ Community!

Clement Kao
Clement Kao
Clement Kao is Co-Founder of Product Manager HQ. He was previously a Principal Product Manager at Blend, an enterprise technology company that is inventing a simpler and more transparent consumer lending experience while ensuring broader access for all types of borrowers.