Cohorts make retention comparable
A raw MRR chart can show growth while hiding the quality of that growth. Cohorts solve this by grouping customers by first active revenue month and tracking each group through time.
This makes it easier to compare customers at the same age. A month one retention cell should be compared with other month one cells, not with a newer cohort that has not had time to mature.
Reading the heatmap
Paradigmo colors cohort MRR retention with a simple operating lens:
- Below 100 percent means the cohort has lost net MRR.
- Between 100 and 120 percent means the cohort is stable or moderately expanding.
- Above 120 percent means the cohort is expanding strongly.
The goal is not only to find the green cells. It is to understand which acquisition periods produced durable customers.
Mature cohorts matter
The newest cohort often has no month one or month three data yet. Scoring it too early creates noise.
For health scoring, mature cohorts are more useful because they have had enough time to show onboarding quality, early churn, and expansion behavior.
What to do with the insight
When a cohort performs better, inspect the common traits: acquisition channel, customer size, plan, agency, region, onboarding process, or product package.
When a cohort performs worse, check whether it contains test rows, non-converted accounts, bad imports, or unusually small customers. The data-quality workflow matters as much as the chart.