Retention Analysis

What is Retention Analysis?

Retention analysis measures what percentage of users acquired at a point in time remain active after subsequent periods: day 7, day 30, day 90. If you acquired 1,000 users in March, retention analysis asks: how many are still active in April (day 30 retention)? How many in June (day 90)? Retention reveals whether users perceive product value (they come back) or whether they churn (they leave and don’t return).

Retention is the core measure of product-market fit. Revenue, user growth, and feature adoption all depend on users remaining engaged. A product that acquires 10,000 users monthly but retains 0% will fail, regardless of acquisition efficiency. A product that acquires 100 users monthly and retains 80% will grow. Retention transcends vanity metrics.

Retention Types: Login vs. Engagement vs. Payment

There are multiple retention definitions depending on the business model. Login retention (user logs in at least once in the period) is lenient—a user might login but not engage. Engagement retention (user completes a meaningful action) is stricter and better—it measures whether users find value. Payment retention (user renews their subscription or makes a purchase) is most meaningful for revenue but only applies to paid products.

Choice of retention metric shapes interpretation. A product with 70% login retention but 40% engagement retention suggests users perceive value (they return) but don’t find it compelling (they don’t act). This suggests product opportunity: make value more obvious or compelling.

Retention Curves & Shapes

A retention curve (plotting retention percentage over time) reveals product behavior. A steep drop in the first week suggests onboarding problems—users don’t see value immediately. A cliff at day 7 or day 30 might suggest a usage pattern or a decision point (trial expires, free credits exhaust). A gentle decline suggests satisfied users with natural attrition. An increasing curve (users returning more frequently over time) is rare but suggests network effects or habit formation.

Understanding your curve’s shape informs solutions. A steep early drop requires onboarding investment; a cliff requires identifying what causes the cliff (trial expiration, feature discovery, etc.) and removing it.

Cohort Retention & Trend Analysis

Cohort retention compares retention curves across acquisition cohorts. If March cohorts have better retention than August cohorts, something changed—product quality, onboarding, marketing messaging, seasonal factors. Identifying the change informs improvements: replicate what March did differently.

Monitoring retention trends (is 30-day retention improving or declining over time?) reveals whether recent product changes are working. A feature launch should improve retention; if it doesn’t, the feature didn’t address retention barriers.

Why It Matters for Product People

Retention is the core health metric. Feature launches, marketing campaigns, and pricing changes should all improve retention. If they don’t, they’re not actually serving users. Focusing on retention forces alignment: what is the product trying to achieve? Is it making users more successful? Happier? More productive? Retention answers this.

For executives, retention is predictability. If you know your 30-day and 90-day retention curves, you can predict future revenue from current acquisition. Low retention requires high acquisition spend to maintain growth; high retention allows profitable growth. This makes retention a strategic lever.

Related Concepts

Retention analysis depends on clean definitions of “active user” and “retention period.” Cohort analysis reveals retention variation across segments. Funnel analysis traces the first-use experience that predicts retention. User onboarding directly affects early retention.