Cohorts
Analytics compare three groups of users:
Comparing the agent cohort with the control cohort shows the total impact of agentic messaging. Comparing it with the legacy cohort shows how it performs against the template-based messaging it replaces.
Retention
The retention chart shows the share of users in each cohort who are still active a set number of days after they were created, grouped by when the user was created.Impact
The impact table shows how the agent cohort compares with the legacy and control cohorts on each measure:- Retention: the share of users still active after a set number of days.
- Streaks continued: how often users keep their streaks going.
- Reactivated: how often users come back after a few days without activity.
- per active user: one row for each metric weighted in your objectives.
- Emails per active user and Push notifications per active user: how many messages users receive.
- Unsubscribes per 1,000 emails.
Message engagement
The message engagement table breaks down performance by channel, and by whether messages were sent by agents or by legacy templates:- Open rate: email opens, or push notifications that opened your app.
- Click rate: messages where the user clicked a link.
- Reengage rate: messages after which the user came back to your app.
- 👍 rate and 👎 rate: how users rated your emails.
Message ratings
Users can rate each agentic email with a 👍 or 👎. Ratings tell you how users feel about the messages they receive, and agents learn from them too, sending more of what users like and less of what they don’t.Viewing messages
The Queue page lists agentic messages in two tabs:- Scheduled: messages agents plan to send, with the user, channel and send time.
- Sent: messages that have already been sent.