# How to Find Your North Star Metric (and Break It Down)

> A north star metric only works once you decompose it into input metrics. Use the stress test, six NSM categories and a worked example to choose yours.

- URL: https://missiongrowth.io/blog/north-star-metric
- Published: 2026-08-22 · Updated: 2026-09-24
- Author: Ömer Furkan Aktaş, Founder, Mission Growth
- Publisher: Mission Growth. Company facts: https://missiongrowth.io/llms.txt

A north star metric is the single number that best captures the value your product delivers to customers. Naming one is the easy part. The real work starts after: proving the candidate survives scrutiny, breaking it into input metrics your team can actually move, and catching the tradeoff it could quietly create before it shows up in someone else's report.

You'll run a stress test on your candidate, sort it into one of six categories with real company examples, decompose it into input metrics with a worked example, and check it against a tradeoff before you commit.

In this guide:
- A six-item stress test that filters out vanity metrics
- Six north star metric categories, illustrated with real companies
- How to find your north star metric and decompose it into input metrics you can move
- A two-step check against the tradeoff your metric could hide
- Whether, and how often, a north star metric is supposed to change

## What a north star metric is, and the test that filters out vanity numbers

A north star metric is the single number that best represents the value your product delivers to customers, and it survives a stress test that most candidates fail on their first pass. It isn't the same thing as a KPI, an OKR or an OMTM (one metric that matters), even though the three get used interchangeably in board decks.

A north star metric vs KPI comparison is where most teams trip first, because each of the four sits at a different altitude and answers a different question.

Confusing them is the fastest way to end up with a scoreboard instead of a working north star metric framework.

| | North star metric | KPI | OKR | OMTM |
|---|---|---|---|---|
| Scope | Whole company, one guiding number | One team or channel | A goal, plus the results that prove it | One team, one metric |
| Time horizon | Persistent, revisited when the business changes | Ongoing, tracked continuously | A fixed cycle, usually a quarter | A fixed 2-6 month window |
| Purpose | Proves the product is delivering value | Proves a channel or function is healthy | Proves progress toward a specific goal | Focuses a team's short-term priority |

OMTM comes from the book *Lean Analytics*: it's team-specific and runs on that fixed window, not the persistent, company-wide number a north star metric is.

None of that matters if the candidate itself is weak. Add up the criteria from three widely used frameworks for what makes a good candidate and you get sixteen overlapping items: 3 from one, 5 from a second and 8 from a third.

Collapsed by merging repeats of the same underlying dimension, six non-redundant checks remain. How to find your north star metric starts with running it against all six, not a brainstorm:

1. **Customer value.** It reflects real value delivered to customers, before it reflects company success.
2. **Leading, not lagging.** It moves before the outcome shows up, the way a real leading indicator should.
3. **Within your control.** An action your team takes this week can move it.
4. **Simple and frequent.** You can state it in one sentence and measure it often enough to track weekly.
5. **Funnel-wide.** It reflects growth across the whole product, beyond one team's slice of it.
6. **Resists gaming.** A team could not hit the number without delivering real value.

That last check is the one most candidates fail quietly. A metric can pass the first five and still let a team win on paper while customers get nothing.

::figure{src="/blog/figures/north-star-metric-1.svg" alt="A six-item north star metric checklist: customer value, leading indicator, team control, simplicity, funnel-wide reach and gaming resistance." caption="Six checks collapse sixteen overlapping criteria from three published frameworks into one non-redundant stress test." width="720" height="293"}

## The six categories of north star metric, with real examples

North star metrics fall into six categories: revenue, customer growth, consumption, engagement, growth efficiency and user experience. Picking the right category depends on how your product makes money and how customers use it. Lenny Rachitsky's 2021 survey of more than 40 growth-stage companies found:

- Revenue: ~50%
- Customer growth: ~35%
- Consumption: ~30%
- Engagement: ~30%
- Growth efficiency: ~10%
- User experience: ~10%

Those are independent shares, not one split that adds up to a whole. Plenty of companies' practice blends more than one category rather than picking a single pure lane.

Real north star metric examples make the categories concrete:

| Company | Category | North star metric |
|---|---|---|
| Airbnb | Consumption | Nights booked |
| Netflix | Consumption | Median view hours per month, its third era |
| Spotify | Consumption | Time spent listening (illustrative label, not Spotify's confirmed metric) |
| Dropbox | Engagement, then customer growth | Monthly active users, then paid customer growth after its B2C-to-B2B shift |
| Figma | Customer growth | Market share |
| Uber | Consumption, then customer growth | Weekly trips in 2017, later market share as it moved away from revenue, per Rachitsky's survey |

Two of the six, Dropbox and Uber, moved categories entirely as the business changed. That pattern gets its own section below, because it turns out to be the norm.

Most companies do best with exactly one north star metric. The exceptions pair a quality metric alongside the primary number, the way Superhuman and Slack do, or run genuinely separate product lines that pull in different directions, the way Spotify's music and podcast businesses do.

Revenue is the most common category, but it's the easiest one to pick lazily: it can climb from a single enterprise deal or a price increase while the product itself gets no better. The same trap shows up one level down, in channel reporting. Proving [SEO ROI](https://missiongrowth.io/blog/seo-roi) convincingly means separating revenue movement from a channel's real contribution, not crediting the channel for revenue it didn't cause.

A B2B SaaS company optimizing net-new logos usually belongs in customer growth. New signups decide that category; how deeply existing users engage decides consumption instead. A [SaaS SEO strategy](https://missiongrowth.io/blog/saas-seo) built to reach that same buyer should be judged against that same number.

A channel-specific number is not a substitute for either. [AI search visibility KPIs](https://missiongrowth.io/blog/ai-search-visibility-kpis) tell you whether your content gets cited in AI answers, which is useful, but it's an input to track, not a company-wide north star metric standing in its own category.

## How to decompose your north star metric into input metrics

A north star metric only becomes actionable once you decompose it into L1 input metrics that directly move it and L2 input metrics that move the L1s. Mixpanel's model adds a third layer, L3, for metrics that feed the L2s, but most teams only need the first two day to day.

Here's how the six categories line up with input-metric layers, using illustrative levers for each:

- **Revenue.** L1: expansion revenue rate. L2: trial-to-paid conversion rate.
- **Customer growth.** L1: new signups. L2: top-of-funnel visitor volume.
- **Consumption.** L1: 7-day return rate. L2: feature adoption rate.
- **Engagement.** L1: weekly active teams. L2: onboarding completion rate.
- **Growth efficiency.** L1: CAC payback period. L2: sales cycle length.
- **User experience.** L1: support ticket rate. L2: page load time.

::figure{src="/blog/figures/north-star-metric-2.svg" alt="A matrix table crossing six growth-metric categories against L1 and L2 input-metric layers, with example levers in each cell." caption="One table crosses six categories against L1 and L2 input-metric layers, the reference no single source combines." width="720" height="369"}

The decomposition isn't a diagram exercise, it's arithmetic. Say a B2B SaaS company picks weekly active teams as its north star metric, sitting in the engagement category above. Written as a function of its input metrics, it's new signups, times activation rate, times 7-day return rate:

- 2,000 new signups
- a 40% activation rate, giving 800 activated teams
- a 60% 7-day return rate, giving 480 weekly active teams

Lift activation to 50% and run the same formula: 1,000 activated teams times 60% return gives 600 weekly active teams, a 25% rise from one input alone.

::figure{src="/blog/figures/north-star-metric-4.svg" alt="A worked example comparing a 40% and a 50% activation rate: weekly active teams rises from 480 to 600 while new signups and 7-day return rate stay flat." caption="Lifting activation from 40% to 50% alone raises weekly active teams from 480 to 600, a 25% rise, with signups and return rate held flat." width="720" height="288"}

Run your own numbers the same way and the diagnosis gets specific fast. If weekly active teams falls, the three factors on the right tell you exactly where to look: fewer new signups points at the top of the funnel, a lower activation rate points at onboarding, and a lower 7-day return rate points at retention.

That's the difference between decomposition and a diagram. A diagram shows boxes; the formula tells you which box moved.

Decomposition only works when each input metric has a named owner. If everyone reports to the same top-line number with nobody responsible for the levers underneath it, the north star metric becomes a scoreboard nobody can move.

Those input metrics are also what your [growth experiments](https://missiongrowth.io/blog/growth-experiment-cadence) should target, run on a fixed cadence rather than whenever someone has an idea. And an input metric is frequently the same number a [growth loop](https://missiongrowth.io/blog/growth-loops) is built to move, so once your north star metric is decomposed, that loop gives you the mechanism for compounding it instead of chasing it campaign by campaign.

## Guard against the metric your north star could quietly break

A north star metric can rise while a real tradeoff metric silently falls, so name the metric your candidate could hurt before you lock it in. Brian Balfour, of Reforge, draws the underlying distinction: output metrics are lagging indicators of results, while input metrics are leading indicators of the actions that produce them. A rising output can hide a leading indicator that's already breaking.

Run this two-step check before you commit to a candidate:

1. Write down the metric your candidate could plausibly trade off against.
2. Set a guardrail threshold on that metric, so a rising north star metric triggers a check the moment the guardrail crosses it.

Balfour's own illustrative example shows the mechanism: a company optimizing ad revenue per user on a news feed can boost that number while eroding the long-term retention and engagement that produced it in the first place. This is a stated illustration, not a documented event at a real company.

Casey Winters, Pinterest's former head of growth, describes a real, documented case from 2016. Pinterest once combined two different actions, a repin and a click, into one "weekly active repinner or clicker" metric. Winters attributes the term to Pinterest's head of product: false rigor.

The combined number could rise even when nobody could say which specific behavior was actually driving it. A metric can rise for the wrong reason and still look like a win on the dashboard.

Funnel-stage metrics, the kind [pirate metrics](https://missiongrowth.io/blog/pirate-metrics) tracks stage by stage, are one common source of a tradeoff partner. A north star metric built from one stage of the funnel can rise while an earlier or later stage quietly degrades underneath it.

## Does your north star metric ever change?

Your north star metric is not permanent: about a quarter of growth-stage companies change theirs, usually for one of four specific reasons. growthmethod.com's own comparison table lists its time horizon as persistent, saying it changes rarely.

Lenny Rachitsky's 2021 survey of more than 40 growth-stage companies tells a different story. About a quarter of them, roughly 25%, had recently changed their north star metric or were about to.

Four triggers show up in that survey, each backed by a named company:

1. **Business-model pivot.** Dropbox moved from engagement (measured as monthly active users) to paid customer growth as it shifted from a B2C to a B2B business.
2. **New product line.** Spotify's north star metric refocused toward consumption once it launched its podcasting business.
3. **Market-share phase.** Figma and Uber both moved their north star metric away from revenue toward market share as they matured.
4. **Provable gaming.** If check six from the stress test above starts failing in practice, that's a trigger on its own, independent of any business-model change.

Netflix is the clearest documented case. Per Rachitsky's survey, it moved through three distinct north star metrics as its business model changed: first the percentage of DVDs arriving in the mail by the next day, then the percentage of members watching at least 15 minutes of streaming a month, and now median view hours per month.

::figure{src="/blog/figures/north-star-metric-3.svg" alt="A timeline of Netflix's three guiding metrics: DVD next-day delivery rate, then 15+ minute streaming share, then median view hours per month." caption="Per Rachitsky's survey, Netflix changed its guiding metric three times as its business model shifted from DVD-by-mail to streaming." width="720" height="184"}

A revised north star metric means a revised growth model, too: once it changes, the drivers rolling up into it change with it. [How to build a growth model](https://missiongrowth.io/blog/growth-model) covers the mechanics of turning a decomposed metric into the full quantitative model underneath it.

A north star metric earns its name only when you decompose it, stress-test it, and revisit it deliberately, not when you pin a word on a slide. If you already have a candidate, run it through the six-item stress test this week, write its 2-3 input metrics on the same page, and name the one metric it could be quietly trading against before your next planning cycle starts.

## FAQ

### What is Uber's north star metric?

Weekly trips. In his own words from a 2017 Y Combinator panel, Uber's first VP of growth, Ed Baker, said: "Ultimately we decided on trips, weekly trips as the North Star metric because for every trip to take place you need riders and you need drivers."

### What is Spotify's north star metric?

Widely reported as time spent listening, a framing attributed to former VP of growth Mayur Gupta's comments about prioritizing engagement depth over raw acquisition. Spotify's own documentation offers it only as an illustrative example, not a confirmed internal metric, so treat the label as directional.

### What is Airbnb's north star metric?

Nights booked, decomposed into input metrics including guest conversion rate, new homes listed and site visitor volume, per Lenny Rachitsky's survey of Airbnb's growth practice.

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

Most do best with exactly one. The exceptions pair a quality metric alongside the primary number, the way Superhuman and Slack do, or run genuinely separate product lines, the way Spotify's music and podcast businesses do.

### What are L1 and L2 metrics?

L1 input metrics move your north star metric directly. L2 input metrics sit one layer upstream and move the L1s. Mixpanel's framework adds an L3 layer for metrics that feed the L2s, though most teams only track the first two day to day.

### What is the north star framework?

The north star framework is the full system: the north star metric at the top, the input metrics that feed it below, and the process for reviewing both on a schedule rather than treating either as fixed forever.