How to use this hub
Start with a crisp definition of your retention denominator (logos vs revenue) and your cohort logic, then use the calculators in the sidebar to compute churn/retention, GRR/NRR, and curve shapes. Only after definitions are stable should you set targets.
Metric map (what each one answers)
| Metric | Answers | Best used for |
|---|---|---|
| Churn rate (logos) | How many customers left- | Customer experience and retention risk |
| MRR churn | How much recurring revenue churned- | Revenue impact from churn |
| GRR | How much revenue did we keep before expansion- | Baseline retention health (hard to fake) |
| NRR | Did expansion offset churn- | Growth quality and account expansion |
| Retention curve | Does retention stabilize over time- | Product-market fit and cohort behavior |
A simple workflow (weekly/monthly)
- Pick one primary view per audience: logos for SMB, revenue (GRR/NRR) for mid-market/enterprise.
- Compute churn/retention for a consistent period (monthly is common).
- Layer GRR/NRR to separate churn vs expansion effects.
- Track cohort curves to see whether early churn is improving and whether long-term retention is stable.
- Set targets with a buffer (targets break when definitions or mix changes).
Common mistakes (that break dashboards)
- Mixing cohorts (signup cohorts vs paid cohorts vs activation cohorts) in the same chart.
- Counting expansion as 'retention' without also tracking GRR (NRR can hide churn).
- Changing event definitions or billing rules without versioning (trends become meaningless).
- Ignoring mix shift: enterprise NRR can rise while SMB churn worsens (segment separately).
When to use targets vs curves
Targets are useful for operating cadences, but curves reveal the underlying shape of retention and whether improvements are durable. If you only look at a single churn number, you'll miss whether cohorts are stabilizing or just shifting churn later.