Dark Mode
Credit Card User Churn Analysis
Finance & Banking Analytics
Tags and Keywords
Trusted By
"No reviews yet"
Free
About
A manager at a bank is experiencing a growing number of customers discontinuing their credit card services. This dataset aims to facilitate the prediction of customers who are likely to churn, allowing the bank to proactively engage with them, offer enhanced services, and potentially influence their decision to remain with the bank.
Columns
- CLIENTNUM: A unique identifier for each customer account.
- Attrition_Flag: An internal event variable indicating whether an account is closed (1) or not (0).
- Customer_Age: The customer's age, expressed in years.
- Gender: Indicates the customer's gender, with 'M' for Male and 'F' for Female.
- Dependent_count: The number of dependents associated with the customer.
- Education_Level: The customer's educational qualification, examples include high school or college graduate.
- Marital_Status: The customer's marital status, such as Married, Single, Divorced, or Unknown.
- Income_Category: The annual income bracket of the account holder, for instance, Less than $40K, $40K - $60K, $60K - $80K, $80K-$120K, or > $120K.
- Card_Category: The type of credit card held by the customer, which can be Blue, Silver, Gold, or Platinum.
- Months_on_book: The duration, in months, of the customer's relationship with the bank.
- Total_Relationship_count: The overall number of products the customer holds with the bank.
- Months_Inactive_12_mon: The count of months the customer was inactive over the past 12 months.
- Contacts_Count_12_mon: The number of contacts made with the customer during the last 12 months.
- Credit_Limit: The credit limit assigned to the credit card.
- Total_Revolving_Bal: The total outstanding revolving balance on the credit card.
- Avg_Open_To_Buy: The average available credit line (open to buy) over the preceding 12 months.
- Total_Amt_Chng_Q4_Q1: The percentage change in transaction amount from Quarter 1 to Quarter 4.
- Total_Trans_Amt: The total transaction amount recorded over the last 12 months.
- Total_Trans_Ct: The total count of transactions performed over the last 12 months.
- Total_Ct_Chng_Q4_Q1: The percentage change in transaction count from Quarter 1 to Quarter 4.
- Avg_Utilization_Ratio: The average utilisation ratio of the credit card.
- Naive_Bayes_Classifier_attribution (Two related columns): Attributes derived from a Naive Bayes classifier.
Distribution
This dataset contains records for 10,000 credit card customers and features nearly 18 distinct attributes for each customer. It is formatted as a structured dataset, typically a CSV file, and has a size of 1.51 MB. Of the total customer base, 16.07% are identified as churned customers, while 84% are existing customers. The data is well-formed, with 10.1k valid entries for most columns. Customer ages range from 26 to 73 years, with a mean age of 46.3. The gender distribution shows 53% female and 47% male customers. The customer relationship period with the bank varies from 13 to 56 months, averaging 35.9 months.
Usage
This dataset is ideally suited for:
- Developing and training machine learning models to forecast credit card customer churn.
- Identifying specific customer segments that are at high risk of leaving the bank's services.
- Enabling banks to implement targeted and proactive customer retention initiatives.
- Conducting analytical studies on the factors influencing customer attrition within the banking sector.
Coverage
The dataset encompasses various demographic variables including customer age, gender, number of dependents, education level, marital status, and income category. It also includes product-specific details such as the type of card held, the duration of the customer's relationship with the bank, and the total number of banking products they hold. Furthermore, it covers customer activity metrics, including months of inactivity and contacts made within the last 12 months. Financial attributes like credit limit, total revolving balance, average open-to-buy credit line, and changes in transaction amounts and counts over a quarterly period, alongside the average card utilisation ratio, are also included. While the data pertains to individual customer behaviours, it does not specify geographic regions or a particular time range beyond annual activity metrics.
License
CC0: Public Domain
Who Can Use It
- Bank managers keen to minimise customer attrition and improve retention rates.
- Data scientists and machine learning engineers tasked with building and deploying predictive churn models.
- Business analysts seeking to uncover insights into customer behaviour and loyalty in the financial industry.
- Researchers and academics focused on studying customer dynamics and predictive analytics applications.
Dataset Name Suggestions
- Credit Card Customer Churn Prediction Dataset
- Bank Customer Attrition Data for Retention
- Credit Card User Churn Analysis
- Customer Retention Prediction in Banking
Attributes
Original Data Source: Credit Card User Churn Analysis
QUALITY