Selecting the Right North Star Metric: Essential KPIs Product Manager Guide

Overview

How It Works

Step-by-Step

1. Step 1: Articulate Your Product's Core Value Exchange

   Before evaluating any metric, write down — in plain language — what value your product delivers to customers. Be specific about the outcome, not the feature. Complete this sentence: 'Our product helps [target customer] achieve [specific outcome] by [mechanism].'

   For example, a project management tool might say: 'Our product helps small teams achieve on-time project delivery by making task coordination effortless.' A food delivery app: 'Our product helps busy professionals enjoy restaurant meals at home by removing the friction of ordering and logistics.'

   This value statement becomes your filter for everything that follows. If a candidate metric doesn't connect to this value exchange, it's the wrong North Star — no matter how easy it is to measure.

   Tip: Interview 5-10 recent customers and ask: 'What would you lose if our product disappeared tomorrow?' Their answers reveal the real value exchange better than any internal brainstorm.

2. Step 2: Brainstorm 8-15 Candidate Metrics

   With your value statement clear, generate a wide list of metrics that could quantify that value. Don't filter yet — this is a divergent phase. Include metrics across several categories:

   • Consumption metrics: sessions, time spent, features used
   • Transaction metrics: purchases, bookings, sends
   • Output metrics: items created, tasks completed, messages sent
   • Outcome metrics: goals achieved, problems resolved, milestones hit
   • Adoption metrics: weekly active users, activated accounts

   For each metric, write it with a specific unit and time frame. Not 'engagement' but 'weekly active projects with at least one completed task.' Not 'usage' but 'monthly meals ordered per active customer.'

   Include your team — engineers, designers, data analysts, and customer-facing roles often surface metrics that product managers miss. The more diverse your brainstorm participants, the better your candidate list.

   Tip: Include at least 2-3 metrics that feel uncomfortably close to measuring customer outcomes rather than product usage. These are often the strongest candidates.

3. Step 3: Apply the Six Evaluation Criteria

   Score each candidate metric against six criteria on a 1-5 scale. These criteria are the core of the selection process:

   1. Value Alignment: Does improving this metric mean customers are getting more value? (Not just using more features, but achieving better outcomes.)
   2. Leading Indicator: Does this metric move before revenue and retention improve? A lagging metric like revenue tells you what already happened. A leading metric predicts what will happen.
   3. Actionability: Can your product team directly influence this metric through product changes, experiments, and improvements?
   4. Understandability: Can every person in your company — from engineering to sales to the CEO — immediately understand what this metric means and why it matters?
   5. Measurability: Can you reliably track this metric with your current (or near-term) data infrastructure? A perfect metric you can't measure is useless.
   6. Revenue Correlation: Is there evidence (or strong logical reasoning) that growth in this metric will eventually drive sustainable business growth?

   Create a simple scoring matrix. Be honest — a metric that scores 5 on measurability but 2 on value alignment is a trap.

   Tip: Have different team members score independently before comparing. Disagreements on scores often reveal deeper disagreements about product strategy that need to be resolved.

4. Step 4: Stress-Test Your Top 2-3 Candidates

   Take the 2-3 highest-scoring metrics and subject them to adversarial stress tests. For each candidate, ask:

   • The Perverse Incentive Test: 'If we optimized ruthlessly for this metric, what bad behavior could it encourage?' For example, 'time spent in app' could incentivize addictive dark patterns rather than efficient value delivery.
   • The Ceiling Test: 'Is there a natural ceiling that would make this metric plateau even as the business grows?' Monthly active users, for instance, can plateau in mature markets.
   • The Decomposition Test: 'Can we break this metric into input metrics that different teams can own?' If a metric can't be decomposed, it's hard to make actionable across the organization.
   • The Gut Check: 'If this metric doubled next quarter, would we be confident the business is healthier?' If the answer is ambiguous, it's the wrong metric.

   This step often eliminates candidates that looked strong on paper but fail under scrutiny. It's the difference between kpis product manager teams aspire to track and the one metric that actually drives the right behavior.

   Tip: The perverse incentive test is the most important. If your metric can be gamed in ways that hurt customers, it will be — even unintentionally.

5. Step 5: Validate with Historical Data

   Before committing, backtest your top candidate against historical data. Look for two signals:

   1. Correlation with retention: Cohorts that scored higher on your candidate metric should show better retention over time. If users who had more 'weekly completed projects' retain at 2x the rate of those who didn't, that's strong validation.
   2. Correlation with revenue: Periods where your candidate metric grew should roughly correspond with revenue growth (with some lag). If the metric grew but revenue didn't follow, the metric may not capture real value.

   If you're at an early stage without much data, do qualitative validation instead. Interview your most engaged customers and your churned customers. Map their behavior to your candidate metric. Do power users naturally score high on it? Do churned users score low?

   This step converts your hypothesis into evidence. Skip it, and you risk building your entire strategy on an unvalidated assumption.

   Tip: Even directional data is valuable here. You don't need statistical rigor — you need to confirm that the metric moves in the direction you'd expect when customers are getting real value.

6. Step 6: Define the Metric Precisely

   Once you've selected your North Star Metric, write a precise definition that eliminates ambiguity. Your definition should include:

   • Exact formula: How is it calculated? What's the numerator and denominator (if it's a ratio)?
   • Time frame: Is it measured daily, weekly, monthly?
   • Inclusion/exclusion criteria: Which users count? Do free trial users count? What about internal accounts?
   • Data source: Where does the data come from? Which event or table?

   For example, instead of 'weekly active projects,' define it as: 'The number of unique projects that had at least one task marked as complete by a non-admin team member in a rolling 7-day window, excluding internal test accounts, sourced from the task_completed event in our analytics pipeline.'

   This precision matters because different interpretations of the same metric lead to conflicting dashboards, confused teams, and eroded trust in the number.

   Tip: Write the SQL query (or pseudocode) that calculates the metric. If you can't write the query, the definition isn't precise enough.

7. Step 7: Build Organizational Buy-In

   A North Star Metric only works if the organization rallies around it. Present your selection to leadership and cross-functional partners with a structured narrative:

   1. Start with the customer value statement (Step 1)
   2. Show the candidate metrics you evaluated and the criteria you used (Steps 2-3)
   3. Explain why the winner passed the stress tests that others failed (Step 4)
   4. Present the data validation (Step 5)
   5. Share the precise definition (Step 6)

   Anticipate objections. Sales may worry the metric doesn't map to revenue targets. Engineering may question measurability. Finance may want a metric closer to the P&L. Address each with evidence, not assertions.

   The goal isn't unanimous enthusiasm — it's informed commitment. People don't need to love the metric; they need to understand why it was chosen and agree to orient their work around it. This alignment is what makes the North Star Metric framework powerful — and it begins with a selection process rigorous enough to earn trust.

   Tip: Frame objections as features: 'Yes, this metric doesn't directly measure revenue — that's by design. Revenue is a lagging indicator of the value this metric captures.'

Examples

Best Practices

• Choose a metric that measures value delivered to customers, not value extracted from them. Revenue is an extraction metric; it follows value delivery but doesn't measure it. The best North Star Metrics are ones where growth means customers are genuinely better off.

• Favor rate or frequency metrics over raw counts for mature products. 'Weekly meals ordered per active customer' is more actionable than 'total meals ordered' because it normalizes for user base growth and reveals whether you're delivering more value per person.

• Revisit your North Star Metric selection when your product undergoes a fundamental strategy shift, not on a fixed schedule. A metric chosen for product-market fit exploration may not serve a scaling-stage product — this connects directly to the skill of evolving your North Star across growth stages.

• Document the 'runner-up' metrics and why they were rejected. This prevents the organization from relitigating the decision every quarter and provides useful context when conditions change enough to warrant revisiting the selection.

• Ensure your North Star Metric can be decomposed into 3-5 input metrics that different teams can influence independently. If the metric can't be broken down, it becomes a spectator sport rather than an actionable guide — this is where identifying input metrics becomes essential.

• Test your metric's understandability by explaining it to someone outside your product team — a new hire, a board member, a customer support agent. If they can't immediately grasp what it measures and why it matters, simplify it.

Common Mistakes

Frequently Asked Questions