One North Star, 2–3 Leading Indicators: Product Management Metrics

The best way to measure product management performance is a focused product scorecard: one north star metric, two to three leading indicators the team can influence, and delivery diagnostics to spot constraints. Which metrics matter depends on your product stage, whether you’re proving early value or capturing revenue. The sections below cover definitions, metric categories, selection rules, and how to instrument them correctly.

Metrics, KPIs, and OKRs: what the terms actually mean

Teams often use “metric,” “KPI,” and “OKR” interchangeably, which causes real confusion when it comes time to decide what to track. A metric is any measurable value: page views, signups, response time. A KPI is a metric chosen as a priority indicator of success, tied directly to a business or product goal. An OKR (Objective and Key Result) is a goal-setting structure: the Objective states the direction, the Key Results are the measurable outcomes that prove you got there, often expressed using KPIs.

The distinction between outcome and output trips up even experienced teams. An output is something you shipped: a redesigned checkout flow, a new onboarding sequence, three new API endpoints. An outcome is the change in customer or business behavior that resulted from it. Shipping is an operational test, not evidence of improved product performance, a point SVPG makes directly: releasing a feature proves nothing until you observe how customers actually use it and whether it delivered the intended benefit.

Before adopting any measure as a KPI, run it through three checks:

• Actionable: if the number moves, does the team know what to do differently?

• Controllable: can the team’s decisions plausibly influence it, or is it driven mostly by external factors?

• Tied to priority: does it map to a stated business or product priority, rather than something that’s just easy to track?

A quick example: an Objective might be “Make onboarding effortless for new teams.” A Key Result could be “Raise activation rate from 40% to 55% this quarter.” The KPI tracked weekly is activation rate itself, while raw metrics like signups per day or time spent on the onboarding page feed into understanding why activation moves.

The three-level metric hierarchy and the product scorecard

A defensible scorecard separates measures into three layers, according to SVPG’s product scorecard framework: business outcomes, product outcomes, and delivery diagnostics. Business outcomes are the company-level results: revenue, retention, market share. Product outcomes are the customer behavior changes your product work is meant to produce: activation, engagement depth, feature adoption. Delivery diagnostics measure how reliably and quickly your engineering process ships and stabilizes changes.

The mistake most teams make is building dashboards full of activity metrics, story points completed, features shipped, tickets closed, and calling that a scorecard. Those are output measures. They tell you the team is busy, not that customers are better off.

Building a scorecard that avoids this trap follows a repeatable sequence:

1. Start with the business priority. Pick the one or two outcomes the company most needs this year, such as revenue retention or new market expansion.

2. Choose product outcomes that plausibly drive that priority. If retention is the business priority, activation rate and feature-adoption depth are strong product-outcome candidates.

3. Add delivery diagnostics as a check, not a goal. These monitor whether your engineering process can support the pace of product outcome work.

4. Limit priority KPIs to five or six. SVPG recommends keeping the priority list this small so the team can actually hold all of them in mind; everything else becomes background diagnostic data.

DORA metrics (deployment frequency, lead time for changes, change failure rate, and mean time to recovery) belong firmly in the diagnostic layer. They tell you whether your delivery pipeline is a bottleneck, but they say nothing about whether customers are better served, so they should never replace an outcome KPI.

Pro Tip: Write each KPI on your scorecard next to the one business priority it serves. If you can’t name that priority in five words, cut the KPI.

Core metrics by category, with formulas and use cases

Most product teams need a working vocabulary across five metric categories, then judgment about which ones matter for their stage and business model.

Customer-value metrics measure whether the product is worth what customers pay for it. Net revenue retention (NRR) is calculated as (starting recurring revenue plus expansion minus contraction minus churn) divided by starting recurring revenue, expressed as a percentage. Gross revenue retention strips out expansion, showing only how much revenue you’d keep from a cohort with no upsells at all, according to Scrum Alliance’s KPI guidance for product management. Customer lifetime value (CLV or LTV) estimates the total revenue expected from a customer over the relationship, typically average revenue per account multiplied by expected customer lifespan. These metrics matter most once you have a paying customer base large enough to see patterns, and LiveSession’s revenue analytics guide walks through how SaaS teams typically build these calculations into recurring reporting.

Adoption and engagement metrics show whether people who sign up actually get value. Activation rate is the share of new users who complete a defined qualifying action within a set window, such as creating their first project within seven days. Time to first value (TTFV) measures how long it takes a new user to reach that same qualifying moment. Daily active users over monthly active users (DAU/MAU) is a stickiness ratio: a result near 50% suggests near-daily habitual use, while a result near 10% suggests occasional use, and LiveSession’s DAU/MAU/WAU guide breaks down how to calculate and interpret each variant. Feature-adoption depth tracks what share of active users engage with a specific feature over a period, useful for judging whether a shipped feature earned its place on the roadmap; tactics for lifting this number are covered in more detail elsewhere.

Growth and acquisition metrics answer whether you’re bringing in the right customers efficiently. Customer acquisition cost (CAC) divides total sales and marketing spend by the number of new customers acquired in a period. Cost per acquisition (CPA) is a narrower version often applied to a single channel or campaign. Funnel conversion rate measures the share of users who move from one funnel stage to the next, such as trial signup to paid conversion. A partner analysis of content-driven SaaS growth offers a useful outside view on how acquisition channels get evaluated against these numbers.

Revenue and monetization metrics matter most once the product has proven initial value and the business shifts focus to capture. Monthly recurring revenue (MRR) and annual recurring revenue (ARR) are the predictable revenue baseline from subscriptions. Average revenue per account (ARPU) divides total recurring revenue by the number of active accounts. Expansion revenue captures the additional recurring revenue from existing customers upgrading or adding seats. Churn rate, the share of customers or revenue lost in a period, is calculated as customers lost divided by customers at the start of the period, and LiveSession’s churn calculation guide walks through cohort-based versions of this formula in more depth. A broader SaaS metrics overview covers how these figures interact across a subscription business.

Operational quality metrics, such as defect density (bugs per unit of code or per release) and support ticket volume per active user, matter most when quality problems are actively eroding trust or retention. They’re diagnostic rather than headline KPIs for most teams, useful for catching regressions before they show up in churn numbers.

• Customer-value: NRR, gross revenue retention, CLV

• Adoption: activation rate, TTFV, DAU/MAU, feature-adoption depth

• Growth: CAC, CPA, funnel conversion rate

• Revenue: MRR/ARR, ARPU, expansion revenue, churn rate

• Operational: defect density, support ticket volume

Using DORA metrics to diagnose delivery, not to score outcomes

The four DORA metrics measure how well your engineering process ships and stabilizes software, and they’re diagnostics, not proof that customers are better off.

• Deployment frequency counts how often code reaches production; higher frequency usually signals smaller, safer batches of change.

• Lead time for changes measures the time from a commit being made to it running in production, showing how much friction sits in your release process.

• Change failure rate is the share of deployments that cause a failure requiring a fix, rollback, or hotfix.

• Mean time to recovery measures how long it takes to restore service after a failure, a direct indicator of operational resilience.

The 2024 DORA Accelerate report recommends pairing velocity measures with stability measures rather than optimizing either alone. The same research noted a correlation between increased AI adoption and measured decreases in delivery throughput and stability within their sample, a reminder to interpret delivery metrics in context rather than assuming any single input always improves them.

For product managers, the practical risk is treating deployment frequency as a goal in itself. A team can ship constantly while shipping things nobody uses, or ship fast while destabilizing production. Pairing DORA metrics with an outcome KPI, like activation rate or retention, keeps delivery speed honest: fast releases that don’t move the outcome metric are a signal to look harder at what’s actually being built, not just how quickly.

Which metrics matter at each product stage

Metric priorities shift as a product moves through its lifecycle, and SVPG’s stage framework identifies three broad phases worth planning around.

1. MVP stage: the priority is evidence that the product solves a real problem. Focus on a core activation event (did the user complete the thing that proves value?) and early retention, plus qualitative signals like referenceable customers willing to vouch for the product.

2. Growth stage: the priority shifts to acquiring the right users efficiently and proving they stick. Watch acquisition quality (are new signups converting at similar rates to your best channels?), funnel conversion rates, and week-two retention as an early durability signal before full cohort retention curves mature.

3. Value capture stage: the priority becomes monetization. Net revenue retention, expansion revenue, and pricing-sensitivity signals (which plan tiers convert best, where discounting erodes margin) move to the top of the scorecard.

A short checklist for picking three to six KPIs for the next quarter:

• Name the one business priority this quarter serves.

• Pick a core activation or retention metric that reflects real value delivered.

• Add one growth or monetization metric matched to your current stage.

• Add one delivery diagnostic to catch capacity or stability constraints.

• Drop anything that duplicates another KPI’s signal.

How to pick KPIs your team can actually act on

Picking KPIs is less about finding the “right” metric and more about applying consistent decision rules. Four checks matter most: does the metric tie to a stated priority, can the team’s work plausibly control it, will a change in it trigger a clear action, and can you actually observe it reliably with your current instrumentation.

Practitioner guidance from PMRead offers a simple operational test: if a KPI moves significantly this week, the team should know what to do differently. If a change in the number wouldn’t change any decision, it’s not a KPI, it’s just a number on a dashboard.

• Limit primary KPIs to three to six per team, with everything else treated as a guardrail or diagnostic.

• Choose leading indicators with a plausible causal link to the north star metric, not just a correlated one.

• Set guardrail metrics (like error rate or support volume) that must not regress while you push the primary KPIs.

• Review primary KPIs weekly, and escalate to leadership when a guardrail breaches its threshold for more than a week.

The same PMRead analysis describes most dashboards as “anxiety machines”: too many numbers, no clear owner, no obvious next step. A single north star metric with two or three leading indicators avoids that trap because everyone on the team can hold the full picture in their head.

Pro Tip: Before adding a new KPI to the scorecard, write down the decision it would change. If you can’t fill in that sentence, leave it off.

Getting instrumentation and data quality right

Metrics are only as reliable as the events behind them, and vague event definitions are the most common source of bad product data. “User signed up” sounds simple until you have to decide whether it fires on email verification, first login, or account creation, and each choice changes every downstream metric that depends on it. Scrum Alliance’s KPI guidance recommends defining the qualifying event and observation window explicitly before calculating any activation or retention figure.

Retention is a clear example of how definitions change results. A simple cohort retention formula: take the number of users from a signup cohort still active in week four, divide by the total size of that cohort. If ambiguity creeps into what counts as “active” (a login versus a meaningful action), the number becomes uncomparable month over month.

Good data quality practice includes:

• Event contracts: document what each tracked event means and who owns changing its definition.

• Alerts on volume changes: a sudden drop in event volume usually means broken tracking, not a real user shift.

• Sampling awareness: know whether your analytics tool samples data at high traffic volumes, which can distort smaller cohorts.

• Change labeling: tag dashboards when a metric’s definition changes so past and future numbers aren’t compared incorrectly.

When a new question comes up that existing telemetry can’t answer, SVPG’s “Outcomes are hard” piece makes the case plainly: define the problem you’re trying to solve first, then decide whether it’s worth investing in new instrumentation or whether an existing proxy metric is good enough for now. Not every question justifies a new event.

Building dashboards that tell a clear story

A team scorecard and an executive dashboard serve different audiences and should look different. The team scorecard needs the north star metric, two or three leading indicators, and guardrail metrics that flag when something is breaking, all visible at a glance during a weekly review. The executive dashboard needs fewer, more aggregated numbers: overall retention, revenue growth, and a one-line insight explaining what changed and why, rather than raw activity counts.

• Keep the team scorecard to a single screen; if it scrolls, it has too many metrics.

• Pair every chart with a one-sentence interpretation, not just a trend line.

• Use consistent time windows (weekly for teams, monthly or quarterly for executives) so numbers are comparable across reports.

• Flag guardrail breaches visually, separate from primary KPI trends, so they don’t get lost.

The biggest failure mode is what practitioners sometimes call an anxiety dashboard: dozens of metrics moving in different directions with no clear next step attached to any of them. A useful reporting cadence pairs a weekly team check-in on leading indicators with a monthly rollup for stakeholders that connects metric movement to a specific decision or experiment, keeping the narrative tied to action rather than just observation.

A fair way to measure product manager performance

Evaluating a PM fairly starts with a context check: what parts of the outcome can this person’s decisions actually influence, given team size, technical constraints, and how long they’ve owned the area? Judging a PM on company-wide revenue when they own one feature area sets an unfair bar.

Quantitative measures should map to funded product work the PM actually shaped: did the KPIs tied to their roadmap move in the intended direction, and did they ship diagnostics and evidence of real usage rather than just feature counts. Qualitative measures of craft matter just as much: the quality of discovery work behind a decision, how well the PM manages stakeholder disagreement, and whether the team around them is more capable because of their leadership.

• Evidence: what KPIs moved, and what discovery or user evidence backed key decisions.

• Context: what constraints (team capacity, technical debt, market shifts) shaped the results.

• Development plan: one or two specific skills or practices to build next quarter.

A short review template built on these three sections keeps the conversation grounded in what actually happened rather than vague impressions of busyness.

How session-based evidence closes the gap in your metrics

Quantitative metrics tell you activation dropped 12 points last month; they rarely tell you why. Session replay and heatmaps fill that gap by showing exactly where users hesitate, misclick, or abandon a flow before reaching the activation event you’ve defined. Watching a handful of failed sessions next to a cohort’s retention curve often reveals a root cause that a dashboard alone would never surface, like a confusing button label or a broken step on one browser.

LiveSession’s clickmap tools let teams see exactly where attention and clicks concentrate on a page, useful for validating whether a redesigned onboarding step is actually guiding users toward the qualifying event or just adding friction. Pairing this qualitative layer with quantitative cohort data, rather than relying on either alone, is consistent with academic research showing perceptual and behavioral measures need triangulation for reliable product decisions. Focus on GDPR and CCPA compliance and integrations with popular tools means session-level detail can sit alongside the support and CRM data teams already track.

The traps that quietly wreck good metrics programs

The biggest trap isn’t picking the wrong metric, it’s picking too many. A scorecard with fifteen KPIs produces no clear priority and no accountability, because nobody can hold that many numbers in their head at once. A close second is the output fetish: celebrating features shipped or story points closed as if they were proof of value, when they’re just proof of activity. The third trap is tracking metrics the team has no real influence over, like total company revenue for a team that owns one onboarding flow, which breeds cynicism rather than focus.

Three steps fix most of this. Pick one north star metric tied to real customer value. Define your core activation event precisely, in writing, before you calculate anything from it. Instrument cohorts so you can see retention and adoption trends over time, not just point-in-time snapshots. As the product matures, expect the scorecard itself to change: what matters at MVP stage rarely matters the same way once you’re capturing revenue.

Where to go for deeper reading

Three sources are worth bookmarking if you want to go past this overview. SVPG’s product scorecard explains the three-layer hierarchy in the founders’ own words and is the clearest primary source on why output metrics fail as scorecards. The DORA Accelerate State of DevOps report is the authoritative source on the four delivery metrics and how they correlate with organizational performance. Scrum Alliance’s KPI guidance offers practical formulas and definitions for the customer-value and adoption metrics covered above, a good reference to return to when defining a new event.

Turn your metrics into evidence you can act on

Picking the right KPIs only gets you halfway; you still need to know why a number moved, and that’s where most teams get stuck staring at a dashboard with no clear next step. Session replay, heatmaps, and conversion funnels can be paired with existing usage data, so when activation or retention shifts, teams can watch the actual sessions behind the number instead of guessing at the cause.

That combination matters most in the moments covered above: validating whether a redesigned activation flow works, spotting where a funnel leaks, or checking whether a feature you shipped is actually being used the way you intended. LiveSession’s guide to combining usage analytics with session-level insight walks through how product teams typically set this up.

Plans start with a free tier, with Basic at $54 per year and Pro at $83 per year for teams that need more session volume and features; Enterprise pricing is available on request. If you’re ready to see how session evidence fits alongside the scorecard you just built, check LiveSession’s plans or start with a demo.

Sources

• What are KPIs in product management? - Scrum Alliance

• 2024 DORA Accelerate / State of DevOps report - Google Cloud

• Product manager KPI guidance - PMRead

FAQ

What are the five most important metrics for product performance?

There’s no single universal list, since the right metrics depend on your product’s stage, but a working set typically includes one north star metric, activation rate, retention, a growth or acquisition metric like CAC, and a monetization metric like net revenue retention. SVPG’s scorecard framework recommends keeping priority KPIs to five or six total rather than tracking every metric available.

What are the KPIs for product management?

Product management KPIs typically fall into customer-value, adoption, growth, and revenue categories, such as activation rate, feature-adoption depth, customer acquisition cost, and net revenue retention. The specific KPIs chosen should tie directly to the current business priority and the team’s ability to influence them, per Scrum Alliance’s guidance.

What are metrics in product management?

A metric is any measurable value related to product or business performance, such as signups, page views, or session length. A KPI is a metric elevated to priority status because it’s tied to a specific business or product outcome the team is actively trying to move.

How is a north star metric different from a KPI?

A north star metric is the single measure that best represents the core value your product delivers to customers, while KPIs, including leading indicators, are the smaller set of metrics that support or explain movement in that north star. Most teams pair one north star metric with two or three leading indicators, per practitioner guidance on building actionable scorecards.

Why shouldn’t DORA metrics replace outcome KPIs?

DORA metrics like deployment frequency and change failure rate measure delivery speed and stability, not whether customers are getting more value from the product. The 2024 DORA Accelerate report recommends pairing these delivery diagnostics with outcome metrics like retention or activation so fast shipping doesn’t get mistaken for real progress.