Customer_Churn_Analysis_prediction

Case Study Question: Customer Churn Analysis and Prediction for an E-commerce Platform

Question: An e-commerce platform wants to understand the reasons behind customer churn and develop a predictive model to identify customers at risk of leaving the platform. As a data scientist, your task is to analyze the customer data and build a churn prediction model. Your goal is to provide insights into customer behaviors and demographics that contribute to churn, as well as develop actionable strategies to reduce churn and improve customer retention.

Data Source: For this case study, you will use a dataset from a fictional e-commerce platform. The dataset contains information on customer transactions, interactions, and churn status. Below is a brief description of the dataset:

• CustomerID: Unique identifier for each customer.
• Churn: Indicates whether a customer churned (1) or not (0).
• Tenure: Duration of the customer's association with the e-commerce platform.
• PreferredLoginDevice: Preferred device used for logging into the platform.
• CityTier: Classification of cities based on tiers.
• WarehouseToHome: Distance between the customer's home and the nearest warehouse.
• PreferredPaymentMode: Customer's preferred mode of payment.
• Gender: Gender of the consumer.
• HourSpendOnApp: Number of hours spent by the customer on the e-commerce platform.
• NumberOfDeviceRegistered: Total number of devices registered by a user.
• PreferedOrderCat: Preferred product category of the customer in the last month.
• SatisfactionScore: Customer's satisfaction score.
• MaritalStatus: Customer's marital status.
• NumberOfAddress: Number of addresses added by the customer.
• Complain: Indicates whether a customer filed a complaint (1) or not (0).
• OrderAmountHikeFromlastYear: Percentage increase in order amount compared to the last year.
• CouponUsed: Total number of coupons used by the customer.
• OrderCount: Total number of orders placed by the customer in the last month.
• DaySinceLastOrder: Number of days since the customer's last purchase.
• CashbackAmount: Amount of cashback received by the customer in the last month.

Your Tasks:

1. Perform exploratory data analysis (EDA) to understand the distribution of features, detect any patterns, and explore correlations with churn.

2. Clean the data, handle missing values, and preprocess the features as necessary.

3. Identify the proportion of churned customers in the dataset and compare churn rates across different customer segments (e.g., age, gender, location).

4. Visualize the relationships between churn and other factors (e.g., total_orders, total_spent) using plots and graphs.

5. Build a churn prediction model using machine learning algorithms such as logistic regression, decision trees, or random forests.

6. Evaluate the model's performance using appropriate metrics like accuracy, precision, recall, F1-score, and ROC-AUC.

7. Interpret the model's predictions to understand the factors contributing to churn and identify key drivers of churn.

8. Develop actionable strategies to reduce churn based on the insights from the analysis and model predictions.

Project Overview

This analysis aims to explore customer churn in an e-commerce platform and its relationships with various aspects of the business, including customer experience. The dataset used for this study is the E-Commerce Customer Churn Analysis dataset.