retention-analyzer

Retention Analyzer

Measure user retention — how many users come back, how often, and who's at risk of churning.

When to Use

• "What's our 7-day retention rate?"
• "How many new users from last month are still active?"
• "Are enterprise users more sticky than free users?"
• "Show me users who haven't returned in 30 days"
• "What's the churn rate for users who signed up in Q2?"
• "Compare retention between users who completed onboarding vs those who didn't"

Mental Model

Retention is about returning users. There are three common lenses:

1. N-day retention: What percentage of users who were active on Day 0 came back on Day N? (e.g., Day 1, Day 7, Day 30)
2. Cohort retention: Of users who signed up/started in a period, what percentage are still active N days later?
3. Stickiness: How many days per week/month do active users engage? (DAU/MAU ratio)

Fullstory doesn't have a native retention report, so you build it from segments and metrics.

Workflow

Step 1: Define the retention window

Clarify what the user wants:

• N-day return: "Are users active on Day 7 and Day 30?"
• Ongoing engagement: "How many sessions per week do active users have?"
• Churn detection: "Which users haven't returned in 30+ days?"

Step 2: Build a retention cohort

For N-day retention from signup:

fullstory:build_segment("users with first_seen between July 1 and July 7") → cohort_segment

For ongoing activity:

fullstory:build_segment("users with last_seen in last 7 days") → active_users
fullstory:build_segment("users with last_seen before last 30 days") → churned_users

Step 3: Measure return rate

Build a metric that counts users who were active in both the cohort window AND the return window:

fullstory:build_metric(
  query="users who were active between July 1-7 AND active between Aug 1-7",
  output_type="single_number"
)

Compare to the total cohort size to get the retention rate: "1,200 of 5,000 July Week 1 users were active in August (24% 30-day retention)."