Product Management Metrics Glossary: 40+ Definitions for PMs
Complete glossary of product management metrics with definitions, formulas, benchmarks, and links to free calculators
A reference for product management metrics, formulas, and benchmarks.
How to Use This Glossary
The metrics are divided into five categories: SaaS economics, growth, prioritization, experimentation, and execution. Each entry includes a definition, a formula, interpretation guidance, and a reason to investigate further. Links to free PM Toolkit calculators are provided where available.
Use this page to look up a formula or review a metric before a board meeting, quarterly review, or stakeholder discussion.
Published comparisons apply to the samples and periods they describe. This glossary distinguishes observed figures from operating targets and practical guidance. Match the metric definition, period, and customer group before comparing results. The SaaS benchmarks guide includes source details and comparison tables.
SaaS Economics Metrics
SaaS economics metrics help assess the revenue, retention, and acquisition costs of a subscription business.
LTV (Customer Lifetime Value)
Definition: An estimate of the revenue or gross profit a customer will generate over their relationship with the business. State which version you use.
Formula: Revenue LTV: Monthly ARPU / Monthly Customer Churn Rate. Gross-profit LTV: Monthly ARPU x Gross Margin / Monthly Customer Churn Rate. Express margin and churn as decimals. These simple estimates assume stable revenue and churn.
Interpretation: Bessemer uses gross-margin-adjusted LTV:CAC of at least 3:1 as cloud-company operating guidance, not a universal cutoff.1
Check: An estimated LTV below CAC means the chosen value estimate does not cover acquisition cost. Check the assumptions and costs included.
CAC (Customer Acquisition Cost)
Definition: The total cost of acquiring a new customer, including sales and marketing expenses allocated to acquisition, divided by the number of new customers gained.
Formula: Sales and Marketing Spend Allocated to Acquisition / New Customers Acquired
Interpretation: Compare CAC with gross-profit LTV and payback. See the sourced LTV:CAC guidance.
Check: If CAC rises without a corresponding rise in customer value, investigate acquisition efficiency and customer mix.
LTV:CAC Ratio
Definition: The estimated lifetime value of a customer divided by the cost of acquiring them. Use consistent customer groups and gross-margin treatment.
Formula: LTV / CAC
Interpretation: Bessemer uses 3:1 or higher as an operating reference.1 A high ratio alone does not establish underinvestment; check demand, payback, and the estimates.
Check: Below 1:1, estimated customer value is less than acquisition cost.
Calculate your LTV:CAC Ratio →
MRR (Monthly Recurring Revenue)
Definition: Recurring subscription revenue normalized to one month, including the monthly share of annual plans.
Formula: Sum of monthly-normalized recurring subscription amounts
Interpretation: Compare growth using the same period and company size. The SaaS benchmarks guide reports annual growth by ARR band; those figures are not monthly MRR targets.
Check: If MRR is flat while customers are added, inspect expansion, downgrades, churn, and new-customer revenue.
ARR (Annual Recurring Revenue)
Definition: The annualized value of recurring revenue. This is the standard metric for measuring the scale of a SaaS business.
Formula: MRR x 12
Interpretation: Use ARR to compare subscription scale. It is not the same as recognized revenue or cash collected.
Check: ARR growth alone cannot establish runway. Review cash balance, collections, and net cash burn.
ARPU (Average Revenue Per User)
Definition: The average amount of revenue generated per active user over a given period, typically monthly.
Formula: Total Revenue / Total Active Users
Interpretation: Varies widely by segment. Track trends over time rather than comparing across companies.
Check: Declining ARPU often indicates pricing pressure or a shift toward lower-value customer segments.
ARPA (Average Revenue Per Account)
Definition: Similar to ARPU but measured at the account level rather than the individual user level. More common in B2B contexts where a single account may have multiple users.
Formula: Total Revenue / Total Active Accounts
Interpretation: Compare the same customer groups over time. Higher ARPA may reflect expansion, pricing changes, or a different customer mix.
Check: If ARPA falls while seats grow, inspect plan mix and pricing before attributing the change to lost value.
Churn Rate
Definition: The percentage of starting customers lost during a period. Define loss consistently, such as a cancelled subscription; inactivity is a different measure.
Formula: Customers Lost During Period / Customers at Start of Period x 100
Interpretation: Compare matching periods and customer groups. The ChartMogul comparison reports annual customer retention by ARPA and billing model, not universal monthly churn targets.
Check: Sustained churn compounds. At a constant 7% monthly rate, about 58% of the starting cohort would be lost over a year, not nearly all of it.
Revenue Churn (MRR Churn)
Definition: The percentage of monthly recurring revenue lost due to cancellations and downgrades in a given period.
Formula: (Churned MRR + Contraction MRR) / Starting MRR x 100
Interpretation: Compare gross revenue churn with GRR over the same period. Net revenue churn also includes expansion and can be negative; this gross measure cannot.
Check: Revenue churn above customer churn can reflect larger lost accounts or downgrades. Inspect both rather than assuming only high-value customers left.
Calculate your Revenue Churn →
NRR (Net Revenue Retention)
Definition: The percentage of recurring revenue retained from existing customers over a period, including expansion, contraction, and churn. It shows how revenue from your existing customer base changes.
Formula: (Starting MRR + Expansion MRR - Contraction MRR - Churned MRR) / Starting MRR x 100
Interpretation: Above 100% means expansion exceeded losses in the starting customer cohort. Compare like customer segments and periods.
Check: Below 100%, the starting cohort produces less recurring revenue. New sales would need to offset that loss for total revenue to grow.
GRR (Gross Revenue Retention)
Definition: The percentage of recurring revenue retained from existing customers excluding expansion. This isolates how well you prevent revenue loss.
Formula: (Starting MRR - Contraction MRR - Churned MRR) / Starting MRR x 100
Interpretation: Bessemer’s 2021 cloud-company analysis reported gross retention around 85–90% across ARR ranges, with segment differences.1 This is historical evidence, not a target for every business.
Check: Falling GRR means a larger share of starting revenue is being lost, even if expansion keeps NRR high.
CAC Payback Period
Definition: The estimated number of months of customer gross profit needed to recover acquisition cost.
Formula: CAC / (Monthly Revenue per Customer x Gross Margin); express gross margin as a decimal.
Interpretation: Bessemer’s cloud-company targets are under 12 months for SMB, 18 for mid-market, and 24 for enterprise.1 These are operating targets, not survey medians.
Check: Longer payback ties up cash for longer and increases exposure to churn before recovery.
Calculate your CAC Payback Period →
ROI (Return on Investment)
Definition: The percentage return generated relative to the cost of an investment. Used across product development, marketing campaigns, and strategic initiatives.
Formula: (Gain from Investment - Cost of Investment) / Cost of Investment x 100
Interpretation: Positive ROI means estimated benefits exceed costs over the chosen period. Compare returns alongside risk, cash needs, and confidence in the estimates.
Check: Negative ROI on recurring initiatives signals a need to pivot or cut investment.
Growth and Engagement Metrics
Growth and engagement metrics describe acquisition and product use. Review them alongside retention and revenue before drawing conclusions about future performance.
DAU (Daily Active Users)
Definition: The number of unique users who engage with your product on a given day. The definition of "active" varies by product and should be tied to a meaningful action, not just a login.
Formula: Count of unique users performing a qualifying action in a single day
Interpretation: Absolute numbers depend on market size. Track growth rate and consistency rather than raw count.
Check: Falling DAU with flat MAU means fewer monthly users are active on the measured day. Check several days and the product’s expected use pattern.
MAU (Monthly Active Users)
Definition: The number of unique users who engage with your product at least once during a 30-day period.
Formula: Count of unique users performing a qualifying action in a 30-day window
Interpretation: Compare consistent windows and activity definitions. Registered users and active users answer different questions.
Check: If sign-ups grow faster than MAU, investigate activation, repeat use, and whether the measurement windows match.
DAU/MAU Ratio (Stickiness)
Definition: Daily active users as a share of monthly active users. If you average DAU across the month, state that convention. The ratio does not show that the same people return every day.
Formula: DAU / MAU
Interpretation: Judge it against how often the product is intended to be used. A monthly reporting tool and a daily messaging app need different comparisons.
Check: A falling ratio for a daily-use product is a reason to inspect return patterns, not proof of weak product value.
Calculate your DAU/MAU Ratio →
WAU (Weekly Active Users)
Definition: The number of unique users who engage with your product at least once during a 7-day period. Useful for products with weekly usage patterns such as project management or reporting tools.
Formula: Count of unique users performing a qualifying action in a 7-day window
Interpretation: Compare WAU/MAU over consistent windows for products intended for weekly use.
Check: Investigate sharp changes alongside holidays, releases, acquisition, and tracking changes.
Conversion Rate
Definition: The percentage of users who complete a desired action out of the total number who had the opportunity. Applied to sign-ups, trials, purchases, upgrades, and more.
Formula: Number of Conversions / Total Visitors or Users x 100
Interpretation: Match the action, denominator, channel, and time window before comparing rates. A visit-to-trial rate and a trial-to-paid rate are different metrics.
Check: If conversion falls as traffic grows, inspect channel mix, user eligibility, and the funnel before diagnosing the cause.
Calculate your Conversion Rate →
Retention Rate
Definition: The share of a starting customer or user group that remains at the end of a period. Subscription retention and return-to-product retention use different definitions.
Formula: Customer retention: Starting customers still present at period end / Starting customers x 100. Exclude new customers.
Interpretation: Use the sourced retention comparisons for subscription businesses. Match the cohort, period, and retention definition.
Check: A lower retained share warrants investigation of who leaves and when; the rate alone does not explain why.
Calculate your Retention Rate →
N-Day Retention
Definition: The percentage of users who return to your product exactly N days after their first use. Commonly tracked at Day 1, Day 7, Day 14, and Day 30 to map the retention curve.
Formula: Eligible cohort users active on Day N / Cohort users old enough to reach Day N x 100
Interpretation: Compare cohorts at the same age using the same return event and day-boundary rules. Exact-day retention differs from returning on or after that day.
Check: A steep early drop is a reason to investigate onboarding and expected use frequency. It does not identify the cause.
Analyze your Retention Curves →
NPS (Net Promoter Score)
Definition: A measure of customer loyalty based on a single question: "How likely are you to recommend this product to a friend or colleague?" Responses on a 0-10 scale classify users as Promoters (9-10), Passives (7-8), or Detractors (0-6).
Formula: % Promoters - % Detractors
Interpretation: Userpilot reports average NPS of 35.7 across 229 B2B SaaS companies using its dashboard. Its public page does not state the collection window.2 Use comparable sampling and survey context.
Check: Negative NPS means more respondents scored 0–6 than 9–10. It does not measure how many actively discourage others.
PMF Score (Product-Market Fit Score)
Definition: The name can refer to a survey result or a composite score. PM Toolkit combines multiple signals; its score is different from the Sean Ellis survey percentage.
Formula: Sean Ellis survey: Very disappointed respondents / Total eligible respondents x 100. PM Toolkit’s composite also includes other signals.
Interpretation: The 40% very-disappointed threshold is a survey heuristic, not a validated cutoff for the composite. See the PMF guide for the source and limits.
Check: Do not compare a composite score with the survey’s 40% heuristic or use either as proof of product-market fit.
Activation Rate
Definition: The share of new users who complete a defined action within a chosen window. Check whether that action is associated with later retention.
Formula: Users completing key action / Total new sign-ups x 100
Interpretation: Userpilot reports 37.5% mean activation across 62 B2B SaaS companies using its dashboard.2 Definitions vary by product, and the public page does not give a collection window.
Check: Low activation is a reason to inspect the funnel, user intent, and tracking before blaming onboarding.
Viral Coefficient (K-Factor)
Definition: The average number of new users generated by an existing user through a defined referral cycle.
Formula: Average Invites Sent per User x Invite Conversion Rate
Interpretation: Above 1 means each user generates more than one new user on average in that model. Sustained growth also depends on cycle time, retention, and reachable demand.
Check: A low referral coefficient says little about unpaid channels such as search or direct discovery; it does not mean all growth is paid.
Prioritization Metrics
Prioritization frameworks make the criteria and assumptions behind roadmap decisions easier to compare. Their scores still depend on judgment.
RICE Score
Definition: A framework developed by Intercom for comparing Reach, Impact, Confidence, and Effort estimates.
Formula: (Reach x Impact x Confidence) / Effort
Interpretation: Use RICE to rank items relative to each other rather than targeting an absolute score. The highest-scoring items should align with strategic goals.
Check: Similar scores may be reasonable. Check whether the estimates are supported and whether small input changes alter the ranking.
ICE Score
Definition: A simplified prioritization framework scoring initiatives on Impact, Confidence, and Ease. Faster to apply than RICE, making it suitable for rapid triage.
Formula: PM Toolkit uses Impact x Confidence x Ease, with each rating from 1 to 10. Some ICE variants average the ratings; do not compare scores across different formulas.
Interpretation: Compare items within the same backlog using consistent rating definitions. Recalibrate scales periodically.
Check: Unreviewed ratings can drift. Compare estimates with outcomes and recalibrate where useful.
Weighted Score
Definition: A flexible prioritization method where you define custom criteria, assign weights reflecting strategic importance, and score each feature against every criterion.
Formula: Sum of (Weight x Score) for each criterion
Interpretation: The framework is as good as the criteria you choose. Align weights with current business priorities.
Check: Overlapping or poorly defined criteria can double-count the same benefit and make scoring harder.
WSJF (Weighted Shortest Job First)
Definition: A prioritization method from SAFe (Scaled Agile Framework) that divides the cost of delay by the job duration. Items with the highest WSJF are prioritized first.
Formula: Cost of Delay / Job Duration
Interpretation: Use it to compare the estimated cost of waiting with job duration. Review dependencies and mandatory work alongside the score.
Check: Ignoring duration removes the trade-off the method is designed to show.
Opportunity Score
Definition: A framework based on Outcome-Driven Innovation (ODI) that identifies underserved customer needs by comparing job importance against customer satisfaction with current solutions.
Formula: Importance + max(0, Importance - Satisfaction).3 Use the same scoring method for both inputs.
Interpretation: Higher scores identify important outcomes that current solutions serve poorly. Use the ODI survey and scoring method consistently rather than treating any 1–10 ratings as equivalent.
Check: A score alone does not establish demand for a proposed feature; test whether the solution addresses the unmet outcome.
Impact/Effort Matrix
Definition: A 2x2 grid that plots features or initiatives by their expected impact (high/low) against the effort required (high/low). The four quadrants are Quick Wins, Big Bets, Fill-Ins, and Money Pits.
Formula: Visual placement based on estimated impact and effort scores.
Interpretation: Prioritize Quick Wins (high impact, low effort) first. Invest strategically in Big Bets (high impact, high effort).
Check: Review low-impact, high-effort work carefully, including whether it is mandatory or enables other work.
Build your Impact/Effort Matrix →
Kano Categories
Definition: A model that classifies product features into five categories based on how they affect customer satisfaction: Must-be (expected basics), One-dimensional (more is better), Attractive (delightful surprises), Indifferent (no impact), and Reverse (causes dissatisfaction).
Formula: Classification through paired functional/dysfunctional survey questions and a lookup table.
Interpretation: Ensure all Must-be features are solid before investing in Attractive features.
Check: Indifferent responses suggest limited satisfaction benefit for the surveyed group. Check other requirements before dropping a feature.
Experimentation Metrics
These concepts help you plan an experiment and interpret the uncertainty in its results.
Statistical Significance
Definition: A result meets a chosen threshold for evidence against the no-effect assumption. It does not establish how useful the effect is.
Formula: For a test with significance level alpha, compare its p-value with that threshold, often 0.05.
Interpretation: Choose alpha before testing. A lower threshold reduces false-positive risk under the test’s assumptions but requires more evidence.
Check: Repeatedly stopping a fixed-horizon test when p drops below the threshold raises false-positive risk.
P-Value
Definition: If there were no real effect, how likely would the test be to produce a result at least this extreme? This interpretation depends on the test’s assumptions.
Formula: Calculated through a statistical test (Z-test, T-test, or Chi-squared) based on sample sizes and observed rates.
Interpretation: Below 0.05 is the standard threshold. Below 0.01 provides stronger evidence.
Check: A p-value of 0.04 does not mean a 96% chance the variant is better. Under the no-effect assumption, results this extreme or more extreme would occur 4% of the time.
Confidence Interval
Definition: A range showing the uncertainty around an estimate. A 95% method would cover the true value in about 95% of repeated studies under its assumptions.4
Formula: Observed Effect +/- Margin of Error, where margin of error depends on sample size and variance.
Interpretation: For the corresponding two-sided test, a 95% interval for a difference that excludes zero indicates significance at the 5% level.
Check: A wide interval leaves a broad range of effects compatible with the data. More observations may help narrow it.
Sample Size
Definition: The number of observations (users, sessions, transactions) required to detect a statistically meaningful difference between experiment variants with adequate power.
Formula: Depends on baseline conversion rate, minimum detectable effect, significance level, and power. Calculated using standard formulas or online tools.
Interpretation: Calculate before launching any experiment. Never decide sample size after looking at results.
Check: Too few observations reduce power and leave more uncertainty. A small sample alone does not increase the false-positive rate of a correctly specified test.
Calculate Required Sample Size →
MDE (Minimum Detectable Effect)
Definition: The smallest difference between control and variant that your experiment is designed to reliably detect. Smaller MDEs require larger sample sizes.
Formula: Determined during pre-test planning based on baseline rate, desired power, and significance level.
Interpretation: Choose the smallest effect that would matter for the decision, then check the traffic and time needed to detect it.
Check: A large MDE can miss useful smaller effects; a small MDE may require more traffic than the product can supply.
Power (Statistical Power)
Definition: How likely your test is to detect an improvement of the size you planned for.
Formula: 1 - Beta, where Beta is the probability of a false negative.
Interpretation: 80% is a common planning choice; 90% needs more observations for the same effect and significance level.
Check: Low power makes useful effects easier to miss. An inconclusive result does not prove there is no improvement.
Lift
Definition: The percentage difference in a key metric between the variant and the control in an experiment.
Formula: (Variant Metric - Control Metric) / Control Metric x 100
Interpretation: Judge the lift against its uncertainty, implementation cost, and business value.
Check: A positive estimate can still be too uncertain or too small to justify a change.
Execution Metrics
Execution metrics describe delivery time and throughput. These metrics also include market sizing, which is fundamental to strategic planning.
Sprint Velocity
Definition: The number of story points (or other units of work) a team completes in a single sprint. It's used to forecast capacity and plan future iterations.
Formula: Total Story Points Completed in a Sprint
Interpretation: Use the same team’s recent completed work to forecast capacity, with room for variation. Avoid comparisons between teams’ point scales.
Check: Large swings make a single average less useful. Inspect work size, interruptions, and changes in the team.
Cycle Time
Definition: The elapsed time from when work actively begins on a task to when it is completed and delivered. This measures execution speed within your development process.
Formula: Completion Timestamp - Work Start Timestamp
Interpretation: Compare similar work with consistent start and end points. Review the median and slower cases rather than a universal day target.
Check: Rising cycle time warrants a look at queues, blockers, and work mix; it does not by itself identify technical debt.
Lead Time
Definition: The total elapsed time from when a request is made (ticket created, feature requested) to when the work is delivered to the customer. Lead time includes waiting time plus cycle time.
Formula: Delivery Timestamp - Request Timestamp
Interpretation: The gap between lead time and cycle time shows waiting before work starts when both measures use the same end point.
Check: Long waits relative to active work are a reason to inspect queues and priorities.
Story Points
Definition: A relative unit of measure used in agile development to estimate the size, complexity, and risk of a user story or task. Story points are not hours; they represent relative effort compared to a baseline story.
Formula: Assigned through team estimation (planning poker, T-shirt sizing, or similar techniques).
Interpretation: The specific scale matters less than consistency. Most teams use a Fibonacci-like sequence (1, 2, 3, 5, 8, 13).
Check: Using story points to measure individual developer productivity undermines their purpose and erodes team trust.
TAM (Total Addressable Market)
Definition: The total revenue opportunity available if a product achieved 100% market share. This theoretical maximum represents the full demand for that product or service type.
Formula: Potential customers x Annual revenue per customer is a bottom-up estimate. A top-down estimate starts with a broader market total and narrows it.
Interpretation: Compare the opportunity with the investment and business goals; a large total alone does not establish a viable market.
Check: Check whether the estimate includes buyers or spending the product cannot serve.
SAM (Serviceable Addressable Market)
Definition: The portion of the TAM that your product can realistically serve given its features, geography, distribution model, and current capabilities.
Formula: TAM x Percentage reachable with your business model
Interpretation: Define filters such as geography, segment, and use case. There is no fixed percentage of TAM that every SAM should meet.
Check: If SAM equals TAM, check the assumptions; equality is possible if the product can serve the whole defined market.
SOM (Serviceable Obtainable Market)
Definition: The portion of the SAM that you can realistically capture in the near term, given your current resources, competitive position, and go-to-market strategy.
Formula: SAM x Realistic capture percentage
Interpretation: Build the capture estimate from sales capacity, demand, competition, and a stated time horizon.
Check: A capture percentage without supporting acquisition and delivery assumptions is not a forecast.
Putting It All Together
Related metrics share inputs, but the relationship depends on the formula and the evidence:
- Lower churn raises simple LTV estimates when other inputs stay fixed. It does not directly shorten the gross-profit CAC payback formula above.
- Check whether activation predicts later retention in your cohorts before treating it as a useful signal.
- Shorter cycle time may allow more experiments, but the number of useful results depends on the questions and test designs.
- Prioritization scores make assumptions visible. Review actual outcomes to learn whether those assumptions held.
Start with the metrics relevant to your role and product stage. Use the formulas and calculators to examine how a change in one input affects related results.
Quick Reference Table
| Metric | Category | Key formula | Interpretation |
|---|---|---|---|
| LTV | SaaS economics | Monthly ARPU x Gross Margin / Monthly Churn | Simple gross-profit estimate; assumes stable inputs |
| CAC | SaaS economics | Sales and Marketing / New Customers | Compare with customer value and payback |
| MRR | SaaS economics | Sum of monthly-normalized subscriptions | Compare growth over matching periods |
| Churn rate | SaaS economics | Lost Customers / Starting Customers x 100 | Match cohort and time period |
| NRR | SaaS economics | Starting-cohort ending MRR / Starting MRR x 100 | Above 100% means cohort revenue expanded |
| DAU/MAU | Growth | DAU / MAU | Match expected use frequency |
| Conversion rate | Growth | Conversions / Eligible Users x 100 | Match action and denominator |
| NPS | Growth | % Promoters - % Detractors | Include all valid responses in the percentages |
| RICE | Prioritization | (R x I x C) / E | Relative ranking based on estimates |
| Sample size | Experimentation | Pre-test calculation | Plan before testing |
| Velocity | Execution | Story Points per Sprint | Forecast using the same team’s history |
| TAM | Strategy | Potential Customers x Annual Revenue per Customer | Check the market definition |
Sources
Footnotes
-
Bessemer Venture Partners, Scaling to $100 Million (2021). Cloud-company analysis and operating guidance; not universal thresholds. ↩ ↩2 ↩3 ↩4
-
Userpilot, SaaS Product Metrics Benchmark Report. Different metric samples drawn from customers using Userpilot dashboards; public collection window not stated. Checked September 13, 2026. ↩ ↩2
-
Strategyn, Discover Hidden Market Opportunities and Quantify Your Customers’ Unmet Needs. Original opportunity algorithm and survey scoring explanation. ↩
-
NIST, Confidence Limits for the Mean and The Relationship Between a Test and a Confidence Interval. ↩