Cohorts
A cohort is a group of customers who share a starting point: they first paid in the same month, came from the same source, or live in the same country. The Cohorts card at the bottom of Revenue → Customers follows each group after that first payment, so you can see which ones keep paying and which ones don't.
The card looks at the customers SnowAnalytics has recorded, so it fills in as your history grows. See Customers for the page around it.
Choose what to compare
Four controls sit above the chart.
| Control | Options |
|---|---|
| Group by | First month, First source, First campaign or Country. |
| Metric | Came back %, Lifetime value or Payments. |
| Window | Week or Month. A week is 7 days and a month is 30. |
| Show as | % for a figure per customer, # for the total behind it. |
The metric decides what each cell counts, for the customers in a group during one window after their first payment:
- Came back %: the share who paid at least once.
- Lifetime value: the net revenue they paid, per customer.
- Payments: the number of payments they made, per customer.
Read the results
- Retention curve: lines for the average and for your oldest and newest groups, showing how the metric moves window by window.
- Month 1 and Month 3 cards: the average across all groups in the first and third window after the first payment, with how many customers that covers. They read Week 1 and Week 3 when the window is a week.
- Best group: the group that came back most in its first window. A group needs at least 5 customers to qualify.
- Cohorts table: one row per group with its Customers, then the Start column and one column for each window after it, up to six. The Average row covers every group. Darker cells are higher; hatched cells haven't been reached yet, and a dashed outline marks a window still in progress.
First month groups customers by the month, or week, of their first payment and lists the last 6 months or the last 8 weeks. The other groups cover customers who first paid in the last 180 days, or the last 8 weeks, and list the eight largest.
Export CSV on the table saves what you see as a file.