Pig E. Bank
To better understand the leading causes to customers leaving the bank, Pig E. Bank management has requested analysis of its customers’ demographics to understand them better. These insights will advise management on how to better retain customers..
Tools used
Excel
Powerpoint
Skills required
Data wrangling and subsetting
Consistency checks
Data merging
Deriving variables
Identifying PII data
Grouping and aggregating data
Vlookup
Understanding decision trees
Reporting in Excel
Data available
Initial Exploratory Analysis
Working to understand the tables, data available, and relationships between tables
Identify data issues to be addressed including:
PII data to be removed to protect customer identify
Data inconsistencies such as abbreviations and full words mixed in a column
NULL values present
Less useful ambiguous values in certain columns
Data Checks and Cleaning
Testing the data integrity and cleaning the data as needed
Data issues identified are cleaned.
No data was imputed as it would have had a significant impact on the results with so few records.
The HasCrCard, IsActiveMember, and ExistedFromBank fields were converted from boolean values to text.
(Changing the boolean values to text was done to make reporting more effective and impactful. Typically, I consider the balance between usefulness and efficiency when addressing this.)
Analysis
Provide full analysis of the data to better understand Instacart’s customers and products and address management’s questions
Customer Demographics vs Customer Retention
All available customer attributes have been assessed to understand their impact to customer retention
Decision tree to determine customer retention
According to the customer demographic analysis, this decision tree captures the primary challenges with customer retention
This can then be used to determine if a customer is expected to stay with Pig E. Bank
Customer Retention Decision Tree
Pig E. Bank Final Report and Recommendations
The decision tree to identify customers at risk of leaving the bank can be used to assess each customer per the criteria below:
The customer attributes that have the highest impact on increasing customer retention:
High Customer activity
Under 40 years old
Female
Bank balance over $100,000
Other attributes have shown less impact:
Credit score
Country
Tenure
Number of Products
Having a credit card
Salary