
Dan Patiño
AI Strategy & Innovation at Coderhouse
Data
A North Star metric is the single metric that best captures the value customers get from a product and that, over time, predicts the company’s revenue. It gives every team one shared definition of success, so product, marketing and growth stop optimizing for numbers that pull in different directions.
Analysts are usually the people who make a North Star real: they propose candidates, test whether they actually lead revenue, define the exact calculation and break it down into input metrics teams can move. This guide explains what makes a good North Star, how to choose one and how to connect it to the day-to-day metrics in your dashboards.
What is a North Star metric?
The concept was popularized by product analytics teams and documented in Amplitude’s North Star Framework. According to Amplitude’s guide to good and bad North Star metrics, a strong North Star has three qualities:
It represents the value users get from the product.
It is within product and marketing’s sphere of influence.
It is a leading indicator of revenue.
Familiar examples include something like “weekly active learners” for an education product, “orders delivered on time per month” for a delivery app, or “weekly active teams collaborating” for a workspace tool. Notice that none of them is revenue itself: revenue is the result, the North Star is the customer value that produces it.
North Star metric vs. KPIs
A company has many KPIs: CAC, churn, conversion rate, NPS and so on. The North Star is not a replacement for them. It sits at the top of the tree as the one outcome they should collectively push forward. If you want a refresher on building solid KPIs, see our guide on KPI examples and how data analysts define KPIs that drive decisions.
A useful way to think about it is in three layers:
Business KPI: the value back to the company (revenue, ARR, paid seats).
North Star metric: the value delivered to customers that drives that KPI.
Input metrics: the three to five controllable factors that drive the North Star.
If the North Star rises but revenue does not, there may be a monetization problem. If revenue rises while the North Star falls, the company may be extracting value it is not creating, which rarely lasts.
How analysts choose a North Star metric
1. Identify the core value moment
Start with the action that shows a customer got what they came for: a completed lesson, a delivered order, a published report. Amplitude groups products into three “games” in its article on how to find your product’s North Star: the attention game, the transaction game and the productivity game. Knowing which game you play narrows the candidates.
2. List candidates and define them precisely
For each candidate, write the exact definition: which event, which users, which threshold and which time window. “Active users” is vague; “users who complete at least one lesson in a calendar week” is measurable.
3. Test whether it leads revenue
This is where analysis does the heavy lifting. Compare cohorts of customers with high and low values of the candidate metric and check retention, expansion and revenue over the following months. A good North Star should separate customers who stay and grow from those who churn.
4. Check that teams can influence it
If no team can move the metric through its work, it will not guide decisions. If one team can move it directly with a trick (for example, sending more notifications), it is probably too easy to game.
5. Keep it to one
For a single product and customer base, one North Star with its inputs is the norm. Multiple North Stars recreate the alignment problem the framework is meant to solve.
Breaking it down into input metrics
Inputs are what teams actually work on. A common heuristic is breadth, depth, frequency and efficiency:
Breadth: how many customers get value (for example, activated accounts).
Depth: how much value each customer gets (lessons completed per learner).
Frequency: how often they come back (active weeks per month).
Efficiency: how quickly they reach value (time to first completed lesson).
Analysts then map each input to the funnel steps behind it. A funnel analysis often reveals which input is the real bottleneck, and experiments show whether a change to that input actually moves the North Star.
A simple worked example
Imagine an online learning platform that picks “weekly active learners who complete at least one lesson” as its North Star. The analyst would:
Write the SQL definition: distinct user IDs with a lesson_completed event in each ISO week.
Validate it: learners above the threshold in their first month retain at much higher rates than those below it (based on the platform’s own cohort data).
Define inputs: new learners activated (breadth), lessons per active learner (depth), active weeks per learner (frequency) and days from signup to first lesson (efficiency).
Build one dashboard with the North Star at the top, the four inputs below and the owning team next to each.
Common mistakes
Choosing revenue as the North Star. Revenue is a lagging business outcome, not a measure of customer value.
Picking a vanity metric. Signups or page views can grow while customers get no value.
Skipping validation. A North Star that does not correlate with retention or revenue sends teams in the wrong direction.
Vague definitions. If two dashboards calculate it differently, it stops being a shared truth.
Never revisiting it. When the business model changes, the North Star may need to change too.
FAQ
Can a company have more than one North Star metric?
Large companies with distinct products or customer bases can have one per product line, but each product should have a single North Star with its own inputs.
Who owns the North Star metric?
Leadership owns the decision, but analysts usually own the definition, data quality and reporting, while each input is assigned to a team.
How often should the North Star be reviewed?
It is tracked continuously, often weekly, and its definition is typically revisited only when strategy or the business model shifts.
Keep learning with Coderhouse
Choosing and validating a North Star metric combines SQL, cohort analysis and dashboard design, which are the core skills of the Data Analytics Course. Explore all three programs:
Data Analytics Course — learn SQL, Python and dashboards to define, validate and track the metrics that guide product decisions.
Forward Deployed AI Engineering Course — learn to build AI applications and agents that put business data to work.
GTM Engineer Course (waitlist) — join the waitlist for early access to the program focused on building automated go-to-market systems.
I'm Dan Patiño, head of AI Strategy & Innovation at Coderhouse. My day-to-day work involves merging the tactical management of e-commerce (CRO, Email Marketing and SEO) with the development of disruptive solutions. I specialize in building internal AI-powered apps to automate tasks and boost innovation within the team. I firmly believe that technology is strategy's best ally. To dive deeper into my professional journey, I'll be waiting for you on my LinkedIn profile.