Overview

Retention reports show how many users return and repeat an event after their first interaction with your product, grouped by day, week, or month. Use them to see how long users stay engaged, when drop-off happens, and which cohorts retain best.

Getting started

Go to Analytics in the sidebar. Reports and dashboards live in the same list, select New report, then choose Retention from the dropdown.

📘 Media pending

Screenshot of the New report dropdown hasn't been captured yet.

Selecting events

A retention report is built from two events:

• Anchor event. The event that starts the cohort, for example Sign Up.
• Return condition. The event a user must repeat to count as retained, for example Product Viewed.

For each event:

1. Select an event from the dropdown.
2. Click the filter icon next to the event to add an event-level filter, if needed.

📘 Media pending

Screenshot of the Events panel hasn't been captured yet.

Setting the retention window

The Retention window setting groups cohorts and measures return activity by Daily, Weekly, or Monthly periods.

• Daily. Tracks whether users return each day after their anchor event.
• Weekly. Tracks whether users return each week.
• Monthly. Tracks whether users return each month.

Match this to your product's usage cycle, daily for apps used every day, weekly or monthly for less frequent products.

Next to it, choose a retention mode, which controls how strictly a period counts as a return:

• Classic. The user came back in exactly that window. For a weekly report, Week 1 means they returned between day 7 and day 13. This is the most common mode and the easiest to understand.
• Rolling. The user came back any time within that window. For a weekly report, Week 1 means they returned any time between day 0 and day 13. More lenient, good for measuring habitual usage.
• Unbounded. The user came back any time after that point, ever. For a weekly report, Week 1 means they returned any time after day 7. Good for measuring long-term stickiness.

📘 Media pending

Screenshot of the Retention window setting hasn't been captured yet.

Setting the measuring method

Under Measuring, choose how return activity is quantified:

• Retention rate. The percentage of the original cohort that returned in each period.
• Unique users. The raw number of users who returned in each period.

📘 Media pending

Screenshot of the Measuring setting hasn't been captured yet.

Applying filters (optional)

Open Filter to narrow the report to a subset of users. Filters apply globally across both events and can match on event attributes, user attributes, account, deal, and task attributes, or segments.

📘 Media pending

Screenshot of the Filter panel hasn't been captured yet.

Adding breakdowns (optional)

Open Breakdown to split cohorts into groups by attribute, event property, or segment. You can add up to two breakdowns per report. Each group shows as its own row in the table and its own line in the chart.

📘 Media pending

Screenshot of the Breakdown panel hasn't been captured yet.

Viewing the retention report

Once both events are selected, Intempt generates the retention chart and table.

• Chart. Defaults to a retention curve, a line chart plotting the percentage of users who returned in each period. Switch it to a plain line chart or a summary metric view using the chart type selector above the chart. If breakdowns are applied, each group gets its own colored line.
• Table. A heatmap with one row per cohort, a Size column showing the cohort size, and one column per period. When measuring by retention rate, cells shade darker as the percentage increases. If a breakdown is applied, cohorts are grouped by segment, and each segment's row header has an arrow to expand and collapse its cohorts.

📘 Media pending

Screenshot of the retention report hasn't been captured yet.

Where to go next

• Reports for an overview of all three report types.
• Creating a funnel report to measure step-by-step conversion instead of return behavior.
• Creating an insights report to analyze trends and event volume.
• Building your first dashboard to add this report to a shared board.