A good North Star metric is not clever branding or a pep-talk number for all hands slides. It is an operational lens that clarifies what creates durable value, forces trade-offs into the open, and keeps teams from chasing noise. Pick the right one and you get consistent progress with fewer arguments. Pick the wrong one and you can hit your number while the business quietly decays.
Across dozens of companies, I have watched both outcomes. I have also learned that the strongest North Stars feel almost boring. They are simple enough for a new hire to remember, concrete enough for an engineer to act on, and sticky enough that a board can probe without warping it. That combination rarely happens by accident. It is built with thought, trial, and the kind of honesty you only get from hard metrics tied to actual customer value.
This article outlines a practical approach to defining and operationalizing a North Star metric, informed by the kind of thinking I see from practitioners at (un)Common Logic and from my own years of working in the middle of growth, product, and finance. The examples lean specific. The advice tilts toward decisions you can make this quarter.
A short story of a good number gone wrong
A B2B SaaS company I advised set Monthly Active Users as their North Star metric. The chart pointed up and to the right. Product celebrated higher login counts after they dropped session timeouts. Marketing shifted budget to top-of-funnel content that juiced trials. Six months later, churn jumped from 2.1 percent to 3.4 percent per month. Revenue goals slipped, despite more “active” users.
It took two quarters to unwind what happened. The team was optimizing for the ease of seeing activity, not for the substance of value delivered. A less tidy, more durable North Star would have been “weekly active teams with at least three automated workflows running,” which they later adopted. That change focused work on features customers retained for, prevented the trial pump from dominating, and brought churn back below 2 percent as expansion revenue increased.
The lesson is not that MAU is always bad. The lesson is that your North Star must be the closest behavioral proxy of long-term customer value that your business can reliably influence and measure. The bar is higher than a vanity metric. It demands logical rigor and operational empathy.
A working definition that passes the smell test
A North Star metric is the single measure of customer value that, when it grows sustainably, the business wins. Three clauses in that sentence do the heavy lifting.
- Single measure: this is not a dashboard. It is a scalar that cuts through internal debates. You can use counter-metrics and guardrails, but the North Star must be nameable without a comma. Customer value: it reflects the experience or outcome that customers actually care about, not your internal process steps. Grows sustainably: you can increase it for a short run through channel hacks or promotions. Sustainability demands that it correlates with retention, unit economics, and brand trust over time.
If the definition feels austere, good. Good North Stars dampen politics. They nudge teams to ask, does this move the one measure that matters, and does it do so without breaking the model.
The difference between North Star, KPIs, and OMTM
Three concepts often tangle here:
- Key Performance Indicators are a set of health and performance measures. You will likely have a dozen or more across acquisition, activation, retention, revenue, and cost. One Metric That Matters is a temporary focus for a team or project. It may change by quarter or milestone. It sharpens execution for a period. The North Star metric persists across quarters and ideally years, especially if it captures the job your product does for customers. It survives leadership changes, new feature sets, and market cycles, even if you refine its definition.
The North Star guides. KPIs validate and diagnose. OMTM points the spear for a sprint.
Criteria for a strong North Star
Over time, I have settled on five tests. If a candidate fails two or more, keep looking.
- Ties to retained value. Positive movement correlates with higher net revenue retention or lifetime value over cohorts. Behavioral and leading. It precedes revenue in the user journey and reflects usage or outcomes, not just dollars collected. Measurable at high frequency. You should be able to see directional change at least weekly without heroic data work. Tamper resistant. It resists easy gaming. If a team can pad it without delivering value, it is fragile. Accessible across functions. Every team can see how their work contributes to it, directly or via leading indicators.
That list looks sober on paper. In practice, it saves quarters of regret. At (un)Common Logic, consultants often use a checklist like this as a forcing function before setting channel targets. It draws a line from marketing experiments to durable usage, which protects go-to-market dollars from chasing shallow wins.
Examples by business model, with trade-offs
Getting concrete helps. Here are typical candidates and the snags you will face.
Consumer marketplace
For a two-sided marketplace connecting buyers and sellers, consider “successful transactions per month” with a success threshold like delivered on time and not refunded within 14 days. This metric ties to liquidity, the backbone of marketplaces. It is leading to revenue but not synonymous with it, and it bakes in a quality standard.
Trade-offs:
- It might underweight new category launches with naturally longer delivery cycles. It can be gamed if sellers bundle items to inflate transaction counts. Setting a minimum order value or using GMV per active buyer as a shadow metric can limit this.
Edge case: in thin markets, “buyer-seller match rate within 24 hours” can be more predictive early on, then cede ground to successful transactions as liquidity stabilizes.
B2B SaaS
Good options often involve activated usage by the unit of value. For workflow automation software, “weekly active accounts with 3+ workflows executing 10+ tasks” beats MAU. For messaging platforms, “weekly active teams with 2+ channels having 50+ messages” tends to correlate with stickiness.
Trade-offs:
- Overly strict thresholds hide progress in small customers. Segment by account size and keep separate thresholds for SMB and enterprise. If your primary revenue driver is seat expansion, usage-based North Stars must be paired with seat adoption as a guardrail to avoid over-serving a small set of power users.
Consumer subscription
Consider “paid subscribers with 8+ sessions per month” for a meditation app or “weekly active subscribers who complete 2+ workouts” for fitness content. The idea is paid, retained engagement, not free trials or skimming.
Trade-offs:
- If you rely on annual billing, monthly activity may look flat even when value accrues in streaks. Map the cadence to real usage cycles. Weekly or 28-day windows often work better for habit products.
Ecommerce
Pure revenue feels natural, but it lures you into discount death spirals. A sturdier option is “orders delivered to returning customers” or “first purchases that lead to a second purchase within 60 days.” The second option connects acquisition to expected payback, which improves bid discipline.
Trade-offs:
- Long replenishment cycles blur the 60-day window for categories like furniture or specialty apparel. In those cases, “orders from email subscribers with 2+ past purchases” can act as a proxy for brand health.
Fintech
For consumer credit, “on-time repayments by active borrowers” supports both customer outcomes and portfolio risk. For B2B invoices, “invoices paid within terms through the platform” centers trust and liquidity.
Trade-offs:
- With regulated products, ensure the metric cannot promote riskier customer cohorts. Pair with credit loss rates by cohort and a fair lending review.
Notice the pattern. The best North Star candidates describe a repeated, verifiable customer outcome that links to retention and margins. They are specific enough to be falsifiable, but general enough that teams can rally around them for years.
How to pick yours without overthinking it
Debates about North Stars drag on because teams smuggle strategy fights into metric selection. Pull those apart. Decide the strategic bet separately, then pick the metric that cleanly reflects progress on that bet.
Here is a focused process that fits a month, not a marathon.
- Start from value creation. Write a one sentence answer to this question: what recurring customer outcome, if it happened more often, would create compounding value for both the customer and us. Map value to behavior. Identify the concrete actions that prove the outcome happened, at a cadence that makes sense. Write it as a formula, including a quality threshold. Validate with cohorts. Check 6 to 12 months of cohort data to test whether higher levels of the candidate metric correlate with higher retention or expansion. Stress-test for gaming. Ask every function, how could we inflate this number without creating value. Add quality gates or counter-metrics to prevent those exploits. Commit for two quarters. Announce the metric, lock it for at least two quarters, and set expectations that you will refine thresholds, not rewrite the core concept.
This list is not abstract. In a recent rollout for a logistics platform, this five step path took 23 calendar days, two analyst weeks, and one board meeting.
Choosing the right unit of account
A tricky, often overlooked choice is the unit you measure against: user, account, team, device, store, household. Pick the one that matches the value exchange in your model.
- For SMB SaaS sold by account, account is usually right. Per user often dilutes real adoption when a few users are very active and others are dormant. For consumer apps used in family contexts, household beats user ids that fragment across devices. Use billing or address to proxy household when necessary. For marketplaces with team-based sellers, the seller entity is a better axis than individual listers.
The downstream effects are large. Your instrumentation, data warehouse models, and experimentation guardrails must align to the chosen unit or your metric will wobble.
From North Star to controllable inputs
A North Star organizes attention, but you still need levers. Break the metric down into controllable inputs that teams can influence weekly. The decomposition varies by business, but the logic is common: volume times quality times frequency.
Take the workflow SaaS example: “weekly active accounts with 3+ workflows executing 10+ tasks.”
- Volume: number of accounts trialing each week. Activation rate: percent of new accounts that build at least one workflow within week one. Expansion: percent that reach 3 workflows by week four. Throughput: average tasks executed per workflow per week. Reliability: percent of tasks executed successfully.
Each component has an owner, an experimentation plan, and a weekly readout. If the North Star is flat, the input breakdown tells you where to dig. This keeps the big number from feeling like a scoreboard without a playbook.
Guardrails and counter-metrics that prevent Goodhart’s law
Every singular metric invites creative misinterpretation. Counter-metrics act like bumpers in a bowling lane. They do not replace the North Star, they keep it honest.
For a marketplace optimizing successful transactions per month, three guardrails work:
- Order refund rate stays below a threshold, often 3 to 5 percent depending on category. New seller onboarding NPS remains within a band, preserving supply growth. Support tickets per 100 orders decline over time, indicating scale without chaos.
For SaaS optimizing activated accounts, pair with product qualified leads quality and net revenue retention. If teams chase low quality trials, these counters will reveal the trade-off quickly.
Instrumentation matters more than wordsmithing
I have seen companies spend weeks debating labels and an afternoon sketching event tracking. Flip that ratio. Without reliable measurement, your North Star becomes folklore.
Practical steps:
- Define events and properties that map exactly to the metric, including quality thresholds. Avoid ambiguous triggers like page views. Favor completed actions with success flags. Log the unit of account on every event. Retrofits are painful. Create a single canonical query in your warehouse that computes the metric. Don’t let every team roll its own. Layer a leading indicator panel for weekly standups that shows the North Star, inputs, and guardrails on the same screen, with the same definitions used downstream in BI tools.
An engineering manager once told me their team spent two sprints to build reliable workflow execution logging. That investment paid back within a quarter, because it removed guesswork from every debate they had about prioritization.
How the metric should evolve across stages
Businesses change. The job your product does may not. A mature North Star adapts by tightening thresholds and clarifying quality, not by reinventing itself every planning cycle.
- Early stage: bias toward activation metrics that predict retention, measured at a short cadence. The goal is signal, not scale. Growth stage: raise quality bars and shift toward sustained usage or success definitions that better match retained value. Scale stage: introduce segment specific thresholds and stronger counter-metrics to protect profitability.
Consider a consumer fitness app. Early, “weekly active users” may suffice. As you mature, tighten to “weekly active subscribers completing 2+ guided sessions.” At scale, segment by cohort age and add “percentage of subscribers maintaining an 8 week streak” as a quality gate. The core idea, delivered workouts lead to ongoing value, stays intact. The precision increases.
Communicating the North Star so people actually use it
Rollout matters. A number buried in a planning doc dies quickly. Give it a name, not a slogan, and teach people how to use it.
- Tell the story of why it matters. Share the cohort analysis that links the metric to retention or LTV. People trust evidence over mandate. Show what it is not. Name the tempting proxies you rejected and why. Offer team specific examples. For support, how ticket deflection efforts impact the North Star through reliability. For finance, how forecasting uses it to estimate revenue trajectory.
I like to run one workshop per function where teams rewrite their OKRs or roadmaps with the North Star and inputs in view. Within two weeks, you will hear the vocabulary in standups. Within a quarter, you will see it in pull request descriptions and sales decks.
Pitfalls I still see, and how to sidestep them
Three themes repeat.
First, mistaking brand reach for value creation. Companies choose total registered users, newsletter subscribers, or app installs. These numbers make marketing happy and product anxious. Replace with behavior that proves use, not exposure.
Second, ignoring base rates. If your category’s natural usage cadence is monthly, a weekly activation bar will label healthy customers as failing. Align measurement windows with real customer rhythms.
Third, letting the metric drift in dark corners of the stack. I have seen at least five flavors of “active user” across a single codebase. Standardize definitions in code and SQL. Add metric tests to your CI pipeline that validate event flows after releases.
When one North Star is not enough
Multi sided products sometimes need layer specific gauges. That does not mean two North Stars. It means one system level North Star and, where necessary, a mirrored sub metric on each side.
A rideshare platform can set “trips completed within ETA” as the North Star. Underneath, driver side activation and rider side frequency become controllable sub metrics. Keep the system lens in the lead to prevent side specific optimizations from breaking the loop.
If you operate multiple distinct products or lines of business, use one North Star per product, nested under a portfolio level financial measure like contribution margin or free cash flow. Tie incentives accordingly to avoid intra portfolio cannibalization.
A few real vignettes
A language learning app pinned its North Star to daily streaks. Engagement looked stellar. Renewal rates lagged. Analysis showed users gamed streaks with low effort sessions that did not correlate with progress or retention. The team shifted to “weekly users completing 3+ lessons at or above level mastery” and added spaced repetition features. Twelve months later, streak counts https://cashmiql027.bearsfanteamshop.com/the-un-common-logic-framework-for-creative-testing were lower, but renewal rose by 9 to 12 percent, and customer reviews began to cite actual improvement.
An SMB expense management platform set “receipts uploaded per month” as its focus. Growth teams mailed scanners to top prospects, which bumped uploads but not paid conversions. Finance pushed back hard. The product team reframed the metric to “active companies with 2+ policy rules enforcing reimbursements within 5 days.” That forced automation work and better admin tools. Within two quarters, net revenue retention climbed from 102 percent to 111 percent, even as top-of-funnel slowed during budget season.
A nonprofit donation platform debated GMV as its North Star. They eventually chose “repeat donors within 12 months” to center donor trust and long term fundraising effectiveness. That pivot shifted content and CRM investments from campaign peaks to stewardship. Year two, repeat donor rate improved by 7 points and donor service tickets fell sharply.
Setting targets without sandbagging
Once you have a metric, you will be asked for a target. Calibrate it with a blend of historical trend, input funnels, and unit economics.
- Start with cohort analysis. If last year’s cohort showed a 20 percent 4 week activation rate, and your roadmap can plausibly move two input levers by a combined 20 to 30 percent, a target of 24 to 26 percent is realistic. Cross check with channel capacity. If your activation plan depends on 30 percent more trials and your paid channels can only deliver 15 percent more at constant CAC, your target is fiction. Tie to financial bounds. If the North Star improves but contribution margin per unit declines, you are creating fragile growth. Set paired guardrails on unit economics to catch this.
The most credible targets are slightly uncomfortable and backed by a math narrative that anyone can replay on a whiteboard. They survive budget reviews and do not crater morale when missed by a hair.
Governance and cadence that keep the metric alive
Cadence turns intent into habit. The best teams establish a regular rhythm around the North Star.
- Weekly: review the North Star, input metrics, and guardrails in a single 30 minute forum with cross functional leads. Focus on deltas and decisions. Monthly: reconcile reported numbers with instrumentation bugs, correct drift, and confirm model assumptions still hold. Quarterly: evaluate whether thresholds need tightening and whether the metric still best reflects the strategy. Change slowly. Announce clearly.
When a metric becomes part of how people talk, not just what they report, it starts to shape culture. The opposite is also true. If the North Star only appears in board decks, it will never escape the spreadsheet.
What (un)Common Logic gets right about North Stars
The name hints at it. Strong metrics depend on un-common logic, the kind that resists easy stories and faces the math. Three moves I see from practitioners at (un)Common Logic are worth copying.
They anchor on cohort outcomes instead of snapshots. Before endorsing a North Star, they stress test it against retention curves, payback windows, and marginal CAC. If the curve bends the wrong way, they throw it out.
They prioritize definitional integrity. A single canonical query, shared by finance, product, and marketing, is the standard. This discipline avoids feuds and lets experimentation compound.
They integrate the metric into execution. Channel plans, product specs, and sales enablement tie directly to inputs that roll up to the North Star. There is line of sight from a search ad to a retained action. That alignment turns the metric into forward motion rather than a poster on a wall.
You do not need outside help to apply this logic, but you do need the courage to pick a measure that will sometimes make you look worse in the short run. That honesty is the point.
A compact checklist for your candidate North Star
- Does higher performance on this metric predict retention or lifetime value across cohorts you care about. Can you measure it weekly, with a clear definition and reliable instrumentation. Is it tied to concrete customer behavior or outcomes, not just revenue booked. Can every function influence it through clear input metrics within a quarter. Have you identified the top two ways to game it and added guardrails to prevent them.
If you can answer yes to all five, you have something you can carry into planning with confidence.
A 90 day adoption plan that works in the real world
- Weeks 1 to 2: pick the candidate, define events, and draft the canonical query. Validate correlations on past cohorts. Run the gaming exercise. Weeks 3 to 4: build the dashboard with the North Star, inputs, and guardrails. Run function workshops to rewrite OKRs with inputs in frame. Weeks 5 to 8: launch two to three experiments per function targeting inputs. Hold weekly cross functional reviews focused on deltas and blockers. Weeks 9 to 12: tune thresholds, fix instrumentation drift, and memorialize definitions in code and documentation. Share early stories of teams using the metric to make calls.
At day 90, the goal is not perfection. It is a working loop where the number informs decisions, and teams can see their fingerprints on its movement.
The quiet power of the right North Star
When a North Star is well chosen, meetings change texture. Engineers ask for better logging because they know why it matters. Marketers argue less about MQL definitions because the activation target is crisp. Finance debates shift from ideology to math. You feel progress not as a crescendo of projects shipping, but as a steady compounding of a customer behavior that pays you back.
That is the real promise of a North Star. It is un-flashy, almost plain. It sticks because it reflects the grain of your business. Find that grain, define it with un-common logic, and protect it with discipline. The rest of your metrics will snap into place.