PWC Customer Retention Analysis

5 min readJul 12, 2022

Background

In my ongoing internship at PWC the second task is to analyze the Customer retention data.

In order to find out the reason why there is a reduction in revenue there is a need to find out if customers are still interested in the service rendered, are the customers churning if they are then there is a need to find out why and take necessary steps to correct it.

This is the essence of this analysis, I checked for the trend on how people are leaving our product for other products, why they are doing so to give recommendations on how to improve our customer retention.

Data Collection

The data was provided by PWC.

Data Preparation

The first thing I did is to study the data and determine the type of analysis id carry out. I did this using Excel. Below is a preview of the dataset in Excel:

Upon having a quick understanding of the dataset I moved to power BI where I carried out my analysis.

In Power BI I used the get data function to connect to the Excel workbook which I loaded to Power Query Editor where I was able to carry out some transformations.

In the original dataset the values in the Churn column are Yes and No, for better understanding I used the replace value to change the Yes to Churned and No to Stayed.

I did the same for Partner column where I indicated if the customer has a partner o not as against the Yes and No used in the original dataset.

I also used conditional formatting to group the duration into 0–1 year, 1–2 years, 2–3years, 3–4 years... See visual below:

Data Analysis

Below is a step by step of how I analyzed and visualized the data.

The total number of customer is 7, 043 from which 1,869 customers left leaving us with 5, 174 customers.

As against using the calculate function to determine the number of Churn from the total number of customers I simply used filter as seen in the visual below:

Since the Churn column has two values; Churned and Stayed, I simply unchecked the stayed so I can have a unique value for Churned customers.

The total charges summed up to 16.05M but considering the fact that some of this customers left we cant add their charges to the Total revenue. I simply used the calculate function to filter out the total charges of customers that stayed which I will consider our Total revenue. This is shown in the visual below:

I did the same to find out the amount that should have been added to our revenue which amounts to our loss.

Total loss

This data spans across 72 months. I got this by using the maximum number of months our customers had stayed with us. As shown in the visual below:

The visual below show the rate at which our customers left based on the type of contract they signed with us which are Month-to-Month, Two years and One year.

The visual above shows most of our customers signed a Month-to-Month contract with us and most of them are leaving. Meaning there is a need to look into why our top customers are leaving.

The visual above shows how our customers churned based on the how long they've been with us:

most of our new customers left
most of our old customers stayed

The visual above shows how our customers churned based on the mode of payment. This shows:

higher percentage of our customers pay with Electronic check and about 50% of them churned which means there’s a need to find out what is wrong with the Electronic check mode of payment and have it fixed.

2. other mode of payment seem to be working fine but the technical team still needs to look into them.

The visual above shows how our customers churned based on the charges paid by the contract type. This shows:

most of the revenue is coming from the month-to-month contract type
the amount of charges for customers that churned is almost equivalent to that of customers that stayed.