Fictura
Insights & sessions

Sessions

How Fictura groups events into sessions, and what the session numbers on Home and the user page mean.

A session is one visit to the app. The SDK stamps every event with a $session_id and starts a new one after the app has been idle for 30 minutes (the default). Events that arrive without a $session_id, from older SDK versions, are grouped by the same rule: a gap of more than 30 minutes between a user's events starts a new session. The user page marks those sessions inf (inferred).

Sessions are built by a background job from the events table, and the most recent 48 hours are recomputed on every run, so a session that was still open shows its final length shortly after it ends.

Home → Sessions

The card covers sessions that started in the selected window. Days are in your time zone.

NumberMeaning
SessionsSessions started in the window, with a bar per day.
Per userSessions divided by the users who had at least one.
Median lengthThe middle session length; p90 is the length 90% of sessions are shorter than.
Bounce rateShare of sessions with one screen (or none) and under 10 seconds.
Paywall → paidOf the sessions that showed a paywall, the share that also converted in the same session.

A session converted when it contains a trial_started, subscription_started or purchase_completed event (after aliases, so purchase_succeeded counts as purchase_completed). Billing events from the RevenueCat webhook arrive from the server shortly after the purchase, so one that lands up to 10 minutes after the session's last event still counts toward it.

User page → Sessions

One row per session, newest first, up to 50: its number among the user's sessions, when it started, how long it lasted, the first and last screen, how many screens and events it had, and whether it showed a paywall, converted or bounced. Platform and app version are the session's latest values for $platform and $app_version; a session whose events carry no platform takes the user's.

Session numbers in Insights

$session_id is an event property like any other, so it can be used in Insights filters and breakdowns.

On this page