What Is Retention Rate in Amplitude

What Retention Rate Actually Measures

Retention rate is the percentage of users who performed an action on or after day N, having done that same (or similar) action on day 0. It answers whether your product keeps users engaged.

// Track signup event to form your cohort
amplitude.track('User Signed Up', {
  signupSource: 'organic',
  email: user.email,
  timestamp: Date.now()
});

How to Measure Retention in Amplitude

Amplitude's Retention Report automates the entire calculation. You pick the cohort event and retention event; Amplitude builds the matrix.

// If measuring feature-specific retention, track distinct events
amplitude.track('User Signed Up', { source: 'email' });

// Later, track the feature use
amplitude.track('Report Exported', {
  reportId: '12345',
  fileFormat: 'pdf',
  userId: user.id
});

Practical Example: D7 and D30 Retention

Here's a real scenario to put this into practice.

// Fetch retention data programmatically via Amplitude API
const fetch = require('node-fetch');

const response = await fetch('https://api.amplitude.com/2/retention', {
  method: 'POST',
  headers: {
    'Authorization': `Bearer ${API_KEY}`,
    'Content-Type': 'application/json'
  },
  body: JSON.stringify({
    start_date: '20240301',
    end_date: '20240331',
    cohort_event: 'User Signed Up',
    retention_event: 'App Opened'
  })
});

const retentionData = await response.json();
console.log(retentionData.data);

Common Pitfalls

• Picking a low-signal retention event: If you use Page Viewed as retention, every analytics dashboard page counts as a 'return visit'—your retention will be artificially high and meaningless. Pick something that reflects actual value: a core feature use, a conversion, or a deliberate action.
• Ignoring cohort size: A 100% D30 retention looks great until you see the cohort had 2 users. Always check the N column. Small samples are noise—wait until you have at least 100+ users per cohort before trusting the percentage.
• Timezone misalignment: Amplitude uses UTC. If your product serves users in multiple time zones, day 0 boundaries may not match the local 'first day.' Check Settings > Organization Settings > Timezone and adjust if needed.
• Confusing D0 retention with D1: D0 shows same-day returns; D1 is 24 hours later. D1 is the meaningful metric for most products. D0 is only useful for all-day engagement (e.g., web apps users might check multiple times per day).

Wrapping Up

Retention rate is the percentage of users who return after day 0. Amplitude's Retention Report calculates it for you: set the cohort event (usually signup), pick the retention event (usually app open), and read the matrix. Your D7 and D30 retention are leading indicators of product-market fit. If retention is dropping, it's a warning to dig into onboarding, feature adoption, or product satisfaction. If you want to track retention automatically across Amplitude, Mixpanel, PostHog, and Stripe—and benchmark your retention against competitors in your niche—Product Analyst can help.