How to Master Growth Metrics for SaaS
Step-by-step guide to Growth Metrics for SaaS. Includes time estimates, prerequisites, and expert tips.
This practical guide shows SaaS teams how to instrument, calculate, and operationalize the growth metrics that matter. You will move from fragmented numbers to a reliable system that drives acquisition, activation, retention, and expansion with confidence.
Prerequisites
- -Admin access to product analytics (Mixpanel or Amplitude) and a tag manager or codebase to install SDKs
- -API access to your billing platform (Stripe, Chargebee, or Recurly) with test and live keys
- -CRM access (HubSpot or Salesforce) and permission to export deal and lead data
- -A data warehouse (BigQuery, Snowflake, or Redshift) and a BI tool (Looker, Metabase, or Mode)
- -A documented event tracking plan or the ability to create one, including UTM naming conventions
- -Basic SQL skills and familiarity with SaaS revenue concepts like MRR movements and cohort analysis
Write precise definitions for MRR, ARR, ARPU, CAC, payback, logo churn, revenue churn, NRR, GRR, activation, and PQL. Decide whether your primary entity is account or workspace and document how seats, usage, coupons, and annual prepay are treated. Choose one north-star metric aligned with customer value, for example activated accounts in 7 days or weekly active teams. Record everything in a tracking plan that includes events, properties, and allowed values.
Tips
- +Document scope for each metric, for example exclude trials and one-time fees from MRR.
- +Create a glossary with examples so growth, product, and finance compute the same numbers.
Common Mistakes
- -Mixing user-level and account-level definitions in the same metric.
- -Defining ARR as MRR times 12 without accounting for temporary discounts or credits.
Pro Tips
- *Version your metric definitions and SQL in a git repo, require pull requests for changes, and publish a changelog so everyone knows when numbers shift.
- *Compute NRR from a movement table with a revenue watermark and exclude credits and one-time adjustments, then reconcile monthly to billing exports.
- *Use organization_id as the primary key and stitch to user_id for activation analysis so account-level retention aligns with user-level behavior.
- *Track dunning events and recovery rates separately to reduce involuntary churn and report them alongside voluntary cancellations.
- *Build a channel-level payback model using gross-margin-adjusted contribution that includes sales compensation, onboarding, and partner fees.