Retention analysis tells you how often users come back after their first visit. Instead of one number, the retention heatmap is a grid where each cell is the percentage of a cohort that returned in a given later period.
Open your website's dashboard and select the Retention tab to see it. The page has three views, switchable from the toolbar: Retention (the cohort heatmap and curve), LTV (lifetime-value cohorts, when revenue data is present), and Bounce.
Anatomy of the heatmap
- Rows (cohorts) — each row is a group of users who first visited in the same period (day, week, or month, depending on the granularity toggle).
- Users column — the size of each cohort: how many users first visited in that period.
- Columns (intervals) — Day 1 is the next day after the cohort's first visit; Week 1 is one week later, and so on. The interval labels follow the granularity you've chosen.
- Cells (return rate) — the % of users from that cohort who came back during that interval.
The first data column (the cohort size) anchors every row. Every user in a cohort visited on their first day by definition, so the return rates further to the right are the telling ones.
What "good" retention looks like
There's no universal benchmark — it depends on your product. But here are common patterns:
- Content / news sites: high Day 1 retention, sharp drop after Day 7, long tail at Day 30+.
- SaaS products: Day 1 typically 30-50% for healthy products; if Day 7 stays above 20%, you have product-market fit signals.
- E-commerce: Day 30 retention is more relevant than Day 1 — repeat purchases tend to space out.
The cell colour encodes retention strength. Stronger retention shows green, mid-range shows amber, and weak retention shows red. A row that turns pale or red right after the first interval is a churn signal.
Reading the rows top-to-bottom
If the most recent cohorts (top of the table) have stronger retention than older ones, your product is improving. If the opposite, recent product changes may have hurt onboarding.
Cohort definition
A cohort is "users who had their first session in period X". Use the granularity toggle in the toolbar to switch how cohorts and intervals are grouped:
- Daily — one row per day, intervals measured in days.
- Weekly — one row per week, intervals measured in weeks.
- Monthly — one row per month, intervals measured in months.
The date range selector sets how far back the analysis looks. Cohorts that are less than seven days old are excluded, since they haven't had time to show a meaningful return pattern yet.
How the percentages are calculated
For each cohort, Zenovay finds every user's first visit, then checks which of those users returned in each later interval. The cell value is returned ÷ cohort size, rounded to one decimal place. Hover any cell to see the exact returned / total count behind the percentage.
Plan limits
Retention analysis is available on Pro and above. The plan also sets how far back the date-range selector can look:
| Plan | Retention available | Max look-back window |
|---|---|---|
| Free | No | — |
| Pro | Yes | 12 months |
| Scale | Yes | 4 years |
| Enterprise | Yes | Custom (unlimited by default) |