Retention Rate: Complete Definition & Calculation Guide

Retention rate is the percentage of customers, users, employees, or students who remain active with the same organization over a specific time period. It measures a starting group and tracks how many are still active at the end of that period.

Three main types exist:

• Customer retention rate: The share of customers who continue with the same product or service from start to end of the period. “Active” could mean a paid subscription, a purchase, or a login within the window.
• Employee retention rate: The share of employees who stay employed with the same employer throughout the period. It excludes new hires who joined during the period.
• Student retention rate: The share of students who remain enrolled in the same school or program in the next term or academic year.

Retention differs from persistence. Retention means staying with the same organization. Persistence means staying in the category while possibly switching, like a customer who cancels one app and moves to a competitor.

The standard retention rate formula is: Retention Rate = (Customers at End − New Customers) / Customers at Start × 100

This formula expresses the percentage of the starting group that remains by the end of a specific period. Each component means:

• Customers at Start: Count of distinct customers at the beginning of the period
• New Customers: Count of customers acquired during the period who weren’t in the starting group
• Customers at End: Count of distinct customers at the end of the period

The result is a percentage. Multiplying by 100 converts a decimal into percent format.

Here’s how retention rate calculation compares to :

A simple process keeps the math and definitions consistent. This example uses a cohort, so only original members are counted.

1. Define the cohort

Decide exactly who is in the starting group. Examples include first-time customers who signed up in September, all active subscribers on October 1, or users in a specific country on a free plan.

2. Pick the time window

Choose the period you want to measure: 7-day, 30-day, 90-day, quarterly, or annual retention. Keep the same length when comparing different cohorts.

3. Count users still active

Find how many people from the original cohort are active at the end of the window. Active can mean any clear criterion, such as:

• Paid subscription still open
• At least one purchase made
• Key in-product action completed once in the window

Exclude anyone who joined after the start, duplicate accounts, and test or bot traffic.

4. Run the formula

Example: A September cohort contains 800 new users, and 600 of those users are active on day 30. Retention rate = 600 / 800 × 100 = 75%.

A good retention rate varies by industry, business model, and time period measured. It depends on how often users are expected to return and what qualifies as “active” use.

Different sectors have different expectations:

• SaaS (subscription software): Expectations differ for self-serve vs. enterprise products, monthly vs. annual renewals, and team vs. individual workflows
• Ecommerce and retail: Retention often means within 30 or 90 days. Categories with frequent replenishment differ from big-ticket categories
• Mobile apps: Categories use different windows—day 1, day 7, day 30 for games and consumer apps; weekly or monthly for utilities and productivity

Time period matters too. Short windows capture habit formation for new users; longer windows align with billing cycles or purchasing rhythms. Average retention rate varies significantly based on these factors.

Retention makes more efficient. Keeping existing customers often costs less than finding new ones, increases the money earned per customer over time, and stabilizes future revenue.

Key benefits include:

• Cost-effectiveness: Retained customers require less advertising and sales spend to buy again
• Customer lifetime value (CLTV): Higher retention extends average customer lifespan, which raises
• Expansion revenue: Customers who stay longer are more likely to upgrade, add seats, or buy add-ons
• Revenue predictability: Strong retention stabilizes and repeat purchase rates

Measuring retention in digital products relies on event-based . Event-based data captures what users do inside a product and when they do it so that retention can be tied to specific actions, not just visits or page views.

Point solutions like Google Analytics, which summarize traffic or sessions, often report activity without linking it to in-product behavior. Comprehensive analytics platforms connect user identity, events, and time windows to show who returns, when, and after which actions.

Cohort analysis method

A groups users by a shared start moment and tracks how many return over equal time buckets. Rows represent cohorts defined by a start date, and columns represent days, weeks, or months since that start.

Common setup elements include:

• Start event: The first action that defines entry into the cohort (like Sign Up)
• Return event: The action that indicates continued use (like Play Song or Open App)
• Time window: The range over which returns are evaluated (like eight weeks)

N-Day retention tracking

An N-Day retention curve shows the share of a cohort that returns on specific days after the start date. Day 1 (D1), Day 7 (D7), and Day 30 (D30) are standard checkpoints used to monitor early, mid, and later engagement.

Two common definitions exist:

• N-Day retention: Users return on the exact day N
• Unbounded retention: Users return on or after day N

Finding correlations between user actions and retention rates doesn’t prove causation. Two methods help identify whether specific behaviors actually cause better retention:

Hypothesis testing

Hypothesis testing involves an H0 (null hypothesis) and H1 (primary hypothesis). The null hypothesis is the opposite of your primary hypothesis. The primary hypothesis points to the causal relationship you’re researching.

Example hypotheses for a music app:

• H1: If a user joins a community within the product in the first month, they will remain customers for over a year
• H0: There is no relationship between joining a community and user retention

The goal is to reject the null hypothesis with —ideally with a minimum of .

A/B testing experiments

can bring you from correlation to causation. Change one variable to create two versions (variants A and B), and measure what happens.

For the community example: Split 1,000 users into two groups. Require the first half to join a community when they sign up (variant A) and the other half not to (variant B). Run the experiment for 30 days, then compare retention rates between groups.

If users forced to join a community have higher retention rates, you have evidence for a causal relationship between community joining and retention.

Amplitude connects retention measurement with product improvement in a single workflow. The platform includes analytics, experimentation, audience management, and data governance, so teams move from “what happened” to “what to change” without switching tools.

The process works like this: Define clean tracking and identities, set retention and cohorts, analyze behaviors linked to retention using and correlations, form hypotheses and run A/B tests, ship winning variants, and with alerts.