Customer churn is the percentage of customers that stop using a company service or product during a time period. To calculate the churn rate, divide the number of customers the business lost during a time period by the number of customers the business had at the beginning of that time period. Customer churn rate is an important metric for any business to evaluate because it costs more to acquire new customers than it does to retain existing ones. Returning customers are likely to spend more on various company products and services because they do not need convincing to become a customer.

A bank wants to identify their credit card customers who are likely to churn and the factors that affect their decision to leave in order to mitigate the churn and improve their customer retention strategies. The bank has supplied this dataset containing about 10,000 customer records and 20 different variables that were tracked over a period of time. The dataset columns include demographic data such as age, gender, dependents, education level, marital status & income. The dataset also includes customer financial data such as how many bank services/products a customer uses, card category, relationship length, transaction counts & amounts. We also have a target column called the attrition flag which tracks whether a customer record is currently existing or attrited.

We aim to help improve the banks' customer retention strategies by identifying the biggest factors that lead to customer churn. We will do this by analyzing each feature against the target column to determine what features have the strongest impact on the target. The data is structured into several data points that the bank tracks for each customer. We will only use a single dataset for this project.

We obtained this dataset from LEAPS.

The dataset contains 10,127 records. There are 20 features and 1 target (Attrition Flag).

The data is structured and we will use supervised learning models to make the predictions.