Churn is the silent killer of growth—you can't fix what you can't measure. Amplitude's retention analysis helps you identify which users are leaving, when they leave, and what behavior precedes the drop-off. By tracking the right events and cohorts, you can spot churn patterns before they impact your metrics.

Define Your Churn Signal with Event Tracking

Start by deciding what counts as churn in your product. Is it a user who hasn't logged in for 7 days? 30 days? The JavaScript SDK makes it easy to log the events that matter.

Initialize the Amplitude SDK and track user logins

Set up the Amplitude JavaScript SDK in your app and track a login event every time a user authenticates. This becomes your baseline for measuring active users.

javascript

import * as amplitude from '@amplitude/analytics-browser';

// Initialize Amplitude
amplitude.init('YOUR_API_KEY', {
  userId: user.id,
  defaultTracking: {
    sessions: true,
    pageViews: true,
    formInteractions: true
  }
});

// Track login event
amplitude.track('login', {
  loginMethod: 'email',
  environment: 'web'
});

Initialize Amplitude and log user login events to establish baseline activity

Track key engagement events

Beyond login, log events that signal an active, engaged user—feature usage, subscription renewal, content creation. The more specific your engagement markers, the better your churn model.

javascript

// Track key engagement events
amplitude.track('feature_used', {
  featureName: 'advanced_filters',
  section: 'dashboard'
});

amplitude.track('report_created', {
  reportType: 'custom_metrics',
  duration: 120
});

amplitude.track('subscription_renewed', {
  planTier: 'pro',
  renewalValue: 99.99
});

Log specific engagement actions to define what active users look like

Watch out: Don't rely solely on login. A user might log in but never use your product. Track the actions that actually indicate value extraction.

Create a Retention Cohort in Amplitude

Once your events are flowing into Amplitude, use the Retention chart to measure how many users return after their first session. Amplitude will automatically calculate churn as 100% minus retention.

Navigate to Insights and select Retention

In the Amplitude dashboard, click Analytics > Insights, then choose the Retention chart type. This shows you cohorts of users and how many come back over time.

javascript

// Use Amplitude API to fetch retention data programmatically
const cohortData = await fetch('https://api.amplitude.com/api/2/cohort/list', {
  method: 'GET',
  headers: {
    'Authorization': `Bearer ${YOUR_API_KEY}`,
    'Accept': 'application/json'
  }
});

const cohorts = await cohortData.json();
console.log('Active cohorts:', cohorts.data);

Fetch cohort data from Amplitude API to automate churn monitoring

Define your returning behavior

In the Returning section of the Retention chart, select the event that defines an active user coming back. For most SaaS, this is any login event. For engagement-heavy products, pick a specific feature usage event like dashboard_viewed or feature_used.

javascript

// Log a specific event that will be your 'returning user' marker
amplitude.track('dashboard_viewed', {
  dashboardType: 'overview',
  widgets: ['metrics', 'trends'],
  sessionDuration: 45
});

// In Amplitude UI, use this event as the 'Returning' behavior
// Users who trigger dashboard_viewed after day 1 are counted as retained

Track specific events and use them as retention markers in your cohort analysis

Tip: Test different returning behaviors. Your retention curve should flatten after a few weeks. A sharp cliff usually means you're measuring the wrong activity.

Set Alerts and Export Data for Action

Monitoring churn is only useful if you act on it. Use Amplitude's API and alerting features to trigger workflows in your support and product tools.

Enable anomaly detection on retention drops

In your retention chart, enable Anomaly Detection to flag unusual drops in cohort retention. Amplitude will notify you if retention dips below your historical baseline, signaling a potential product issue or churn surge.

javascript

// Use Amplitude REST API to check retention metrics
const retentionCheck = await fetch(
  'https://api.amplitude.com/api/2/retention?cohort_definition_type=event&returning_window=day&event=dashboard_viewed',
  {
    method: 'GET',
    headers: {
      'Authorization': `Bearer ${YOUR_API_KEY}`
    }
  }
);

const retention = await retentionCheck.json();
if (retention.data[0].retention < 0.45) {
  console.warn('Churn spike detected:', retention.data[0]);
  // Trigger alert to your Slack or monitoring tool
}

Query retention via API and programmatically alert when churn spikes

Create a churned-user cohort for re-engagement campaigns

In Cohorts, create a saved cohort of inactive users—those who haven't performed your key event in the last 30 days. Export this list or sync it to your CRM to trigger re-engagement messaging.

javascript

// Tag users for reactivation based on inactivity
amplitude.setUserProperties({
  churn_risk_30d: true,
  lastEngagementDate: lastLoginDate,
  segmentForCampaign: 'reactivation'
});

// Track users entering the churn cohort
amplitude.track('user_churned', {
  reason: 'no_login_30days',
  lastActiveDate: lastLoginDate,
  planTier: userPlan
});

Tag churned users with properties to sync them to your CRM

Tip: Distinguish between churn types. A user who churns on day 1 (never engaged) needs different outreach than a power user gone silent for 60+ days.

Common Pitfalls

Measuring retention without controlling for seasonal patterns. January churn often differs from summer churn. Use Amplitude's Compare feature to look at year-over-year cohorts.
Setting your returning behavior too high (e.g., requiring advanced feature usage) and missing churn in your free-tier users. Start broad, then refine as you understand your user segments.
Confusing day 1 retention with day 7 retention. Day 1 includes new users, but day 7+ is where real product stickiness emerges. Focus your analysis on D7 and beyond.
Ignoring power-user churn. A 5% overall retention drop might hide the fact that your top 10% of users are churning at 30%. Always segment by usage level.

Wrapping Up

You now have a clear path to monitor churn in Amplitude: track engagement events, build retention cohorts, and set up alerts. By watching your D7, D30, and D90 retention curves, you'll catch churn before it becomes a growth crisis. If you want to track this automatically across tools, Product Analyst can help.

How to Monitor Churn Rate in Amplitude

Define Your Churn Signal with Event Tracking

Initialize the Amplitude SDK and track user logins

Track key engagement events

Create a Retention Cohort in Amplitude

Navigate to Insights and select Retention

Define your returning behavior

Set Alerts and Export Data for Action

Enable anomaly detection on retention drops

Create a churned-user cohort for re-engagement campaigns

Common Pitfalls

Wrapping Up