Customer Churn Analysis and Prediction

Analysis to pin down what is wrong with Comcast’s customer service

Dhruval Patel
CodeX
4 min readMay 13, 2022

--

Photo by David Hahn on Unsplash

Welcome to my story! For a variety of service difficulties, telecom consumers can file complaints with service providers via email/call centers/SMS. It’s very important to resolve complaints as it’s a very competitive industry. Here, we’ll analyze the data of Comcast Telecom Consumer Complaints.

Dataset Description

Comcast is an American global telecommunication company. The firm has been providing terrible customer service. They continue to fall short despite repeated promises to improve. Only last month (October 2016) the authority fined them $2.3 million, after receiving over 1000 consumer complaints. The existing database will serve as a repository of public customer complaints filed against Comcast. It will help to pin down what is wrong with Comcast’s customer service.

I started with importing the libraries, then I import the dataset and clean the dataset. I removed the duplicates and changed the data type of the ‘Date’ and ‘Date_month_year’ columns to datetime64. There are no null values, so I am good to go for an Exploratory Data Analysis (EDA).

From EDA, I found that most of the complaints had been received in the month of June, and from Georgia, Comcast had received the most complaints.

Find my Kaggle notebook here.

Statistical Tasks

(1) Provide the trend chart for the number of complaints at monthly and daily granularity levels

Monthly Complaints Trend
Daily Complaints Trend

June Month has the highest number of complaints

(2) Provide a table with the frequency of complaint types

Complaint Catagory
% Complaints Received

Complaints related to Internet are the most (34.36%)

(3) Create a new categorical variable with values of Open and Closed. Open & Pending is to be categorized as Open and Closed & Solved is to be categorized as Closed

Number of Open complaints are 517 and Closed complaints are 1707

(4) Provide state wise status of complaints in a stacked bar chart. Use the categorized variable from Q3

State-wise Complaints

(5) Which state has the maximum complaints

State-wise Complaints

Comcast had received the most complaints from Georgia

(6) Which state has the highest percentage of unresolved complaints

% Complaints

Georgia has the highest % of unresolved complaints

(7) Provide the percentage of complaints resolved till date, which were received through the Internet and customer care calls

50.62% and 49.42% complaints are resolved which was received by customer care call and Internet respectively

I have drawn some insights —

(1) June month has the highest number of complaints November month has the least number of complaints

(2) Complaints related to the Internet are the most

(3) Around 30 % of the complaints are open

(4) Complaints received from Georgia are the most (288/2224)

(5) Georgia has the highest % of unresolved complaints

(6) 50.62% and 49.42% of complaints are resolved which was received by customer care call and the Internet respectively

Thank you for reading! I would appreciate it if you follow me or share this article with someone. Best wishes.

Your support would be awesome❤️

--

--

Dhruval Patel
CodeX
Writer for

I write technical blogs explaining my Data Science project walkthroughs and the concepts relating to Data Science