Learn Building Cohorts in Python

Cohort analysis is a powerful technique in business analytics that helps you understand the behavior and lifecycle of different groups of customers over time. By grouping customers into cohorts based on shared characteristics—such as their acquisition month—you gain valuable insights into how customer engagement, retention, and value evolve. This longitudinal perspective is essential for identifying trends, measuring the effectiveness of marketing strategies, and making informed decisions to improve customer experience and lifetime value.


              12345678910111213141516
            
import pandas as pd

# Simulated e-commerce transactions data
data = {
    "customer_id": [101, 102, 101, 103, 104, 102, 105, 106, 101],
    "order_date": [
        "2023-01-15", "2023-01-20", "2023-02-10", "2023-02-15",
        "2023-03-01", "2023-03-05", "2023-03-15", "2023-04-01", "2023-04-10"
    ],
    "order_id": [5001, 5002, 5003, 5004, 5005, 5006, 5007, 5008, 5009]
}

df = pd.DataFrame(data)
df["order_date"] = pd.to_datetime(df["order_date"])

print(df)


              12345678910111213
            
# Assign each customer to a cohort based on their first purchase month

# Find the first order date for each customer
df["cohort_month"] = (
    df.groupby("customer_id")["order_date"]
    .transform("min")
    .dt.to_period("M")
)

# Add a column for the order month
df["order_month"] = df["order_date"].dt.to_period("M")

print(df[["customer_id", "order_date", "order_id", "cohort_month", "order_month"]])

The logic of cohort assignment is straightforward but highly effective. By determining each customer's first purchase month, you define a cohort month—the period when they began their relationship with your business. This cohort label remains constant for each customer, regardless of when they place subsequent orders. By comparing behaviors across cohorts, you can track how different groups progress, spot changes in retention or engagement, and assess the long-term impact of initiatives. This method enables you to analyze customer journeys over equivalent timeframes, revealing patterns that would be hidden in simple aggregate data.

Everything was clear?

Thanks for your feedback!

Section 1. Chapter 1

Ask AI

Ask anything or try one of the suggested questions to begin our chat

Suggested prompts:

Can you explain how to calculate retention rates using this cohort data?

What are some common visualizations for cohort analysis?

How can I interpret the results of a cohort analysis?

Swipe to show menu


              12345678910111213141516
            
import pandas as pd

# Simulated e-commerce transactions data
data = {
    "customer_id": [101, 102, 101, 103, 104, 102, 105, 106, 101],
    "order_date": [
        "2023-01-15", "2023-01-20", "2023-02-10", "2023-02-15",
        "2023-03-01", "2023-03-05", "2023-03-15", "2023-04-01", "2023-04-10"
    ],
    "order_id": [5001, 5002, 5003, 5004, 5005, 5006, 5007, 5008, 5009]
}

df = pd.DataFrame(data)
df["order_date"] = pd.to_datetime(df["order_date"])

print(df)


              12345678910111213
            
# Assign each customer to a cohort based on their first purchase month

# Find the first order date for each customer
df["cohort_month"] = (
    df.groupby("customer_id")["order_date"]
    .transform("min")
    .dt.to_period("M")
)

# Add a column for the order month
df["order_month"] = df["order_date"].dt.to_period("M")

print(df[["customer_id", "order_date", "order_id", "cohort_month", "order_month"]])

Everything was clear?

Thanks for your feedback!

Section 1. Chapter 1