A north star metric is the one number that captures the value customers get from you and predicts revenue before it shows up. Choose it by testing candidates against three criteria: it reflects customer value, it leads revenue, and your teams can move it. Then break it into three to five input metrics and give each input to exactly one team.
What a north star metric is and isn’t
A north star metric is a single company-level measure of value delivered, usually a count of customers completing the core job your product or service exists to do, within a time window. Its real job is settling arguments. When marketing, product and sales disagree on priorities, the tiebreaker is which option moves the north star more.
It isn’t:
- A goal. “5,000 weekly active teams” is a target on the north star. The metric is the definition; targets change every quarter.
- A dashboard. It’s one number plus the handful of inputs that drive it. A slide with twelve “north star” KPIs means there isn’t one.
- A vanity count. Signups, followers, downloads and page views can climb while nobody gets value.
- A replacement for financials. Revenue, margin and cash still run the business. The north star sits beside them as the leading signal.
- An OKR. OKRs are quarterly bets that move the inputs; the north star persists across quarters. Marketing OKRs that tie to revenue shows how the two connect.
Criteria: customer value, leading and controllable
A candidate that fails any of these three will mislead you.
Customer value. It should rise only when customers are better off. The test: if this number doubled next quarter, would customers be happier, or would you just be counting more activity? “Messages sent” doubles with notification spam. “Conversations that got a reply” doesn’t.
Leading. It should move before revenue and predict it. Check this in your own data: customers or accounts scoring high on the candidate should retain, expand or reorder more than those scoring low. If you can’t show that across a few cohorts, you have a guess, not a north star.
Controllable. Teams should be able to move it within weeks through their own work. NPS is tied to value but slow and indirect. “New accounts completing setup within 14 days” is something product, marketing and customer success can push directly.
The time window matters as much as the event. Match it to how often a healthy customer naturally uses you: daily for messaging, weekly for most work software, monthly for invoicing or payroll, and the replenishment cycle for consumable products. A weekly window on a product people rightly use once a month makes healthy customers look churned.
Before you commit, run each candidate through this list:
- Rises only when customers get the core value, not just when activity increases
- Counts customers or accounts rather than raw events, where possible
- Has a time window that matches natural usage frequency
- Correlates with retention or repeat revenue in your own cohort data
- At least three teams can say how their work moves it
- Anyone in the company can explain it in one sentence
- Has a guardrail metric that would expose gaming
- Can be reported weekly from systems you already have
Building the input metric tree
The north star is an output; nobody can work on it directly. The input tree breaks it into levers teams can act on.
- Write the north star as a formula. Most break into breadth (how many customers), depth (how much value each gets) and frequency (how often). For example: weekly active accounts = newly activated accounts + retained accounts + reactivated accounts.
- Pick three to five first-level inputs. Together they should cover the whole north star with little overlap.
- Break each input one level down into things a team can change. “Newly activated accounts” becomes qualified signups multiplied by activation rate.
- Stop where a team owns a lever. Two or three levels is usually enough. Deeper trees turn into spreadsheets nobody reads.
- Add a guardrail to each input. Name the metric that shows you’re buying growth badly: CAC payback for acquisition inputs, margin for discount-driven inputs, support tickets per account for aggressive feature pushes.
- Validate against history. Pull the last few quarters. When an input moved, did the north star follow? Inputs that never correlate get redefined or dropped.
A hypothetical tree for a workflow SaaS product looks like this:
- North star: accounts completing at least one core workflow per week
- Newly activated accounts: qualified trial signups × activation rate within 14 days
- Retained active accounts: week-4 retention of new accounts, weekly retention of established accounts
- Reactivated accounts: dormant accounts returning within 30 days of a lifecycle campaign
Examples for B2B SaaS, DTC and service businesses
These are starting shapes. Validate any of them against your own cohorts first.
| Business | North star candidate | Main inputs | Guardrail |
|---|---|---|---|
| B2B SaaS, workflow or collaboration tool | Accounts completing a core workflow each week | Newly activated accounts, week-4 retention, active seats per account | Net revenue retention |
| B2B SaaS, usage-priced | Paying accounts above a meaningful monthly usage threshold | Activated accounts, accounts crossing the threshold, usage growth per account | Gross margin per account |
| DTC, consumable products | Customers placing a second order within the replenishment window | First orders from target customers, second-order rate, subscription starts | Contribution margin per order |
| DTC, durable products | New-customer orders delivered and kept past the return window | First orders, return rate, review rate | Contribution margin, discount depth |
| Professional services or agency | Active clients on track against their agreed outcome, reviewed monthly | Qualified intro calls, proposal win rate, on-time milestone delivery | Project margin, team utilization |
| Local or appointment-based service | Completed jobs rated five stars | Booked jobs, completion rate, first-visit resolution rate | Cost per booking, margin per job |
The pattern: SaaS counts accounts doing the core job, which predicts renewal and expansion. DTC counts repeat or kept purchases, because a first order proves only that the ad worked. Service businesses count clients getting the outcome they paid for, which predicts renewals and referrals better than lead volume.
Why revenue makes a poor north star
Revenue is the scoreboard, not the play. It fails the criteria in four ways:
- It lags. Annual contracts and subscriptions mean revenue reflects decisions customers made months ago. When it drops, the cause can be a quarter or more old.
- It measures capture, not value. Price increases, pulled-forward discounts and prepay pushes raise revenue while customer value stays flat or falls.
- Most teams can’t move it directly. Product, support and lifecycle can’t point to how their week changed revenue, so it stops guiding daily decisions.
- It rewards short-term trades. Discounting, over-promising and closing poor-fit customers all lift revenue now and show up later as churn.
Keep revenue, margin and cash as financial targets, and use them as guardrails on the tree. Acquisition inputs in particular need a cost check like CAC payback period, so nobody grows the north star by overspending. Even in DTC, count repeat customers or kept orders rather than revenue, so price changes and promotions can’t disguise what’s happening.
Assigning inputs to teams
A tree without owners is a poster. The rules I use:
- One owning team per input. Other teams can contribute, but one leader answers for the number.
- Owners get the input and its guardrail together. A target on qualified signups with no CAC guardrail invites cheap, poor-fit volume.
- Nobody owns the north star except leadership. Give marketing the north star directly and it will usually buy the fastest-moving component, typically signups.
- Work the tightest constraint first. If activation is weak, more acquisition just fills a leaky bucket.
| Input | Owning team | Typical levers | Guardrail |
|---|---|---|---|
| Qualified signups or first orders | Marketing | Channel mix, offers, landing pages, ICP targeting | CAC payback |
| Activation rate | Product, with marketing on onboarding emails | Setup flow, time to first value, guided onboarding | Support tickets per new account |
| Early retention or second-order rate | Product or retention marketing | Habit loops, replenishment reminders, post-purchase flows | Discount rate |
| Expansion or active seats | Customer success or account management | Adoption reviews, seat invites, upsell timing | Net revenue retention |
| Reactivation | Lifecycle marketing | Win-back sequences, feature announcements | Unsubscribe and complaint rates |
Review the inputs weekly, the north star monthly, and targets quarterly. Setting the north star and wiring every input to an owner is usually the first deliverable in my growth strategy work, because every channel plan after it gets judged against the tree.
Reviewing and changing it as you grow
Review the north star once a year during planning. Between reviews, change it only when a trigger fires:
- The business model changes. Moving from self-serve to sales-led, from one-time purchase to subscription, or adding a marketplace changes where value is delivered.
- The value moment changes. A new core use case or a new ideal customer can make the old definition count the wrong behavior.
- The metric decouples. The north star keeps rising while retention or revenue stops following. That is the clearest sign it’s being gamed or no longer predicts anything.
Stage shifts the emphasis too. Before product-market fit, a company-wide north star is usually premature; watch retention of a small cohort directly. In early growth, count activated and retained customers. At scale, add depth: seats, frequency or repeat orders per customer. With several product lines, each line gets its own north star, all rolling up to shared financial targets.
When you do change it, run the old and new metrics in parallel for a quarter, restate history under the new definition, explain the reason in writing, and rebuild the input tree before resetting team targets.
Get it built
If your teams are working hard on different numbers, I can help you choose the north star, build the input tree from your own data and give every input an owner and a weekly review. Start with a Growth Audit, $1,500 fixed and credited if we continue. See pricing or get in touch.