In Python, load the data with pandas and compute customer retention rate from column totals for the whole period, or row by row for each month. On the sample data it returns 94.75.
Retention rate = (Customers at end − New customers) ÷ Customers at start × 100
What you need
Customer retention rate is the percentage of customers at the start of a period who are still customers at the end: customers at the end minus new customers gained, divided by customers at the start, times 100. You need these columns, here from the sample file monthly.csv:
- df["new_customers"]: new subscribers
- df["customers_start"]: subscribers at start of month
- df["customers_end"]: subscribers at end of month
Step by step
- Install pandas (pip install pandas) and save monthly.csv next to your script.
- Run the script. The first print shows one row per month, the second the figure for the whole period.
- To use your own data, rename your columns to match, or change the names in the code.
The code
import pandas as pd
df = pd.read_csv("monthly.csv")
# Each month
df["retention_rate"] = (df["customers_end"] - df["new_customers"]) / df["customers_start"] * 100
print(df[["month", "retention_rate"]].round(2))
# Whole period: add up the parts first, then divide
retention_rate_total = (df["customers_end"].sum() - df["new_customers"].sum()) / df["customers_start"].sum() * 100
print(round(retention_rate_total, 2))
What you should get
Run against the sample data, the code above returns these values (in percent):
| Month | new_customers | customers_start | customers_end | retention_rate |
|---|---|---|---|---|
| 2026-01 | 120 | 800 | 880 | 95.00% |
| 2026-02 | 135 | 880 | 971 | 95.00% |
| 2026-03 | 150 | 971 | 1,069 | 94.64% |
| 2026-04 | 128 | 1,069 | 1,136 | 94.29% |
| 2026-05 | 160 | 1,136 | 1,239 | 94.98% |
| 2026-06 | 175 | 1,239 | 1,348 | 94.67% |
| Whole period | 868 | 6,095 | 6,643 | 94.75% |
(6,643 − 868) ÷ 6,095 × 100 = 94.75%
Notes for Python
- Dividing by zero in pandas does not raise an error: it gives inf or NaN. Check with df.replace([float("inf"), -float("inf")], float("nan")) before averaging or plotting.
- For the whole period, sum the columns first and then divide. df[col].mean() of the monthly results weights a small month the same as a large one.
- pct_change(), groupby() and resample() extend the same idea to weekly data, product categories and daily rows.
Where customer retention rate misleads
- Take new customers out of the end count, or retention looks higher than it is.
- Reactivated customers (who left and came back in the same period) muddy the count. Decide how to treat them.
- Retention over a year is not monthly retention to the twelfth power if churn changes over the year.
The full explanation, with a calculator, is on the Customer retention rate (CRR) page.
Same metric, other tools
Related metrics in Python (pandas)
Learn more about Python
- Python and SQL for data analysis, side by side — 3 min read
- How to find out what changed in your sales data — 2 min read
About the data on this page
The figures come from a small made-up dataset, printed above in full so you can check every number by hand. Download it as monthly.csv. The results are computed from those rows when the page is built, and the Excel formulas, SQL and Python on these pages are run against the same rows before publishing, so the code and the answers agree.
To get these numbers from your own data without writing formulas, try AnalyzeIt: ask in plain words and it shows the working behind every figure.
Legal entity
Devansh Yadav
Moradabad, Uttar Pradesh
Governing law: Courts at Moradabad, Uttar Pradesh
Phone: +91 8433475698
Legal: legal@analyzeit.in