How to Calculate Retention Rate in Sql

Customer Retention Rate Calculator

Understanding customer retention is critical for the long-term health of any SaaS or subscription-based business. While the calculator above provides a quick way to verify your numbers manually, performing this analysis on large datasets requires Structured Query Language (SQL). This guide explains the mathematics behind the metric and provides the specific SQL queries needed to extract it from your database.

The Retention Formula

Before diving into the code, it is essential to understand the standard retention formula used in business analytics:

Retention Rate = ((E – N) / S) × 100

• E (End): Total number of customers at the end of the time period.
• N (New): New customers acquired specifically during that time period.
• S (Start): Total number of customers at the start of the time period.

Calculating Retention with SQL: Step-by-Step

To calculate retention in SQL, you generally need a table (e.g., subscriptions or users) that records when a customer started and when (or if) they churned.

1. The Simple Snapshot Approach

If you just want a snapshot for a specific month, you can use conditional aggregation (COUNT with CASE statements) to define your S, E, and N variables.

SELECT
  COUNT(DISTINCT CASE WHEN start_date < '2023-10-01' AND (end_date >= '2023-10-01' OR end_date IS NULL) THEN user_id END) AS Start_Count,
  COUNT(DISTINCT CASE WHEN start_date <= '2023-10-31' AND (end_date > '2023-10-31' OR end_date IS NULL) THEN user_id END) AS End_Count,
  COUNT(DISTINCT CASE WHEN start_date BETWEEN '2023-10-01' AND '2023-10-31' THEN user_id END) AS New_Customers
FROM subscriptions;

Once you run this query, you will get the raw numbers ($S, E, N$) which you can plug into the calculator above or wrap in a Common Table Expression (CTE) to calculate the percentage automatically.

2. Cohort Analysis (The "Classic" Retention Query)

A more advanced and common way to view retention is via Cohort Analysis. This tracks a specific group of users who joined in Month 0 and calculates how many are left in Month 1, Month 2, etc.

Here is a standard SQL pattern using a self-join to calculate month-over-month retention:

WITH monthly_activity AS (
  SELECT
    user_id,
    DATE_TRUNC('month', activity_date) AS activity_month
  FROM logins
  GROUP BY 1, 2
)
SELECT
  previous.activity_month AS start_month,
  COUNT(DISTINCT previous.user_id) AS start_users,
  COUNT(DISTINCT current.user_id) AS retained_users,
  (COUNT(DISTINCT current.user_id)::float / COUNT(DISTINCT previous.user_id)) * 100 AS retention_rate
FROM monthly_activity AS previous
LEFT JOIN monthly_activity AS current
  ON previous.user_id = current.user_id
  AND current.activity_month = previous.activity_month + INTERVAL '1 month'
GROUP BY 1
ORDER BY 1;

Common Pitfalls in SQL Retention Calculation

When writing your retention queries, watch out for these common errors that can skew your data:

• Counting Upgrades as New Users: Ensure your query distinguishes between a completely new account and an existing user upgrading their plan. "New Customers" ($N$) must only include genuine acquisitions.
• Null End Dates: In many databases, active users have a NULL value in the end_date or churn_date column. Ensure your logic (IS NULL) accounts for this, otherwise, you will exclude all your active happy customers.
• Timezone Discrepancies: Always standardize your timestamps (e.g., to UTC) before truncating them to months or weeks, otherwise, users active on the last day of the month might be attributed to the wrong period.

Why Use a Calculator?

Even for seasoned Data Analysts, writing SQL queries for retention can be prone to syntax errors. Use the calculator at the top of this page to benchmark your SQL results. If your SQL query says your retention is 95% but the raw counts ($S, E, N$) plugged into the calculator result in 82%, you likely have a logic error in your JOIN or WHERE clauses.