0% found this document useful (0 votes)
24 views24 pages

Telco Customers Churn Predication - Analysis

This case study analyzes customer churn in the telecommunications sector, identifying key factors that lead to customers switching providers. The dataset includes 7043 records and highlights issues such as poor network quality, high pricing, and inadequate customer service as common causes of churn. The analysis reveals that month-to-month contracts have the highest churn rates, while additional services and longer tenures contribute to customer retention.

Uploaded by

lavanyan1205
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
24 views24 pages

Telco Customers Churn Predication - Analysis

This case study analyzes customer churn in the telecommunications sector, identifying key factors that lead to customers switching providers. The dataset includes 7043 records and highlights issues such as poor network quality, high pricing, and inadequate customer service as common causes of churn. The analysis reveals that month-to-month contracts have the highest churn rates, while additional services and longer tenures contribute to customer retention.

Uploaded by

lavanyan1205
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 24

Case Study: Telecommunication Churn

Prediction

Group Members:
Lavanya Miranam
Vibha Chaudhary
Vaishnavi Gaikwad
Ram Khandare

DataSet Link: Telco Customer Churn | Kaggle


What is Customer Churn ?
Customer churn refers to when a customer decides
to stop using a service, and switches to a competitor.
What Is Telcom Customer Churn ?
Customer churn in telecommunication referes the
customers in telcom switching to other service provider
telecom companies, churn is a major concern as losing
customers means losing revenue and market share

Common Causes of Telecom Churn :


 Poor Network Quality (dropped calls, slow internet)
 High Pricing (competitors offering better deals)
 Poor Customer Service (unresolved complaints)
 Lack of Personalization (customers feel undervalued)
 Billing Issues (hidden charges, incorrect bills)
Dataset Information :-
The dataset used in this project was imported from
Kaggle.This dataset is about total 7043 records and 21
variables.
Overview of Dataset:-
Code:- 2. Loading Libraries and Data:-

Here the DataFrame name we set as tel

3.Columns Description:-

 Info about Customer realated columns


: gender,CutomerID,Partner,Dependent,SeniorCitizen(age range)
 Services that each customer has signed up for columns
: 'PhoneService', 'MultipleLines', 'InternetService', 'OnlineSecurity',
'OnlineBackup', 'DeviceProtection', 'TechSupport','StreamingTV',
'StreamingMovies'
 Customers who left within the last month : The Column called
churn
 Customer account information : 'Contract', 'PaperlessBilling',
'PaymentMethod', 'MonthlyCharges', 'TotalCharges',tenure (no of
month they are taking services)

Column Type:-
Code:-
 Numerical Type: tenure,Monthlycharges,TotalCharges.

 Categorical: gender,SeniorCitizen,partner,Dependents,'PhoneService',
'MultipleLines', 'InternetService', 'OnlineSecurity', 'OnlineBackup',
'DeviceProtection', 'TechSupport','StreamingTV',
'StreamingMovies',contract,PaperlessBilling,PaymentMethod,Churn

 Mixed: customerID

Analysis
Issues With Dataset Columns Type :-
 Total Charges must be in form of numerical
 Seniorcitizen column having 0 and 1 it must be yes or
no a categorical type data
Type Conversion:-

Code:-
1. Converting to_numeric

2. Setting No at 0 and Yes at 1


Output:-
Handling Missing Values:-

Here after Transforming TotalCharges object to numeric


we can see there is 11 total missing values
Step1: 1) Check how the Total charges Column is formed
By analysis total charges is nothing but (tenure *
monthly charges + Extra Charges)
From this we understood how to calculate totalcharges

2) Are there any common factors present in other


columns if TotalCharges has NaN values?
Here, we can see that the most common contract type is
"Two-year" and payment method is mailed check when
TotalCharges is NaN and tenure is zero.
To replace Totalcharges nan values we can calculate average
tenure and Extra charges whoes contract type is two year and
payment method is mailed check
Code:-
Create new dataframe:-

Drop tenure =0 records to calculate avg:-


Create New result Column to calculate total charges
TotalCarges=tenure*monthlycharges

Now we can see there is some difference in result and


Totalcarges the difference is nothing but a Extra charges
Create New differe column to calculate Extra Charges
Now Lets fill Tenure=0 records

Here we are filling average of tenure where tenure =0


After that we are filling null values of TotalCharges

Here There is no
missing values in our
dataset

Cleaning Done
Data Visualization
here the most important column is Churn lets
Visualization Churn Rate

Here we can see 26.5% Customers are in Churn


Lets Visualize Churn with all columns
Churn Based On Gender:-
Code

Note: there is no difference in Churn if gender


is different
Most of Churn happened in Month-to-month Contract type

Most of Churn are having Monthly Charges between 60 tp 120


Customers Who are not Depended are in churn

Customers Who are using PhoneService there in churn


We can see
In Internet Service Customers who are using Fier Optic ther
are leaving hence we have to take care or stops service of
fiber optic

Customers who are not taken any online security are in churn
Here also we can see Customers who are not taken
DeviceProtection They are in Churn
Here a Chustomers who are not having techsupport are in
churn

From this technical variable analysis we can conclude that the


customers who are using fiber optical and not taking online
security,tech support,device protection are in churn
And the customers who are not depended them also in churn
Paymenth Method :-

Here most of customers who are having Electronic check they


are in churn by seeing this we can also predict that we have to
focus on electronic billing issue may be happing ?
Numerical Columns Correlation:-
Here tenure and TotalCharges are having highly positive
correlation with each other
Conclusion:-
Customer Churn Patterns:
 A significant number of customers churn in the first few months of
tenure. Customers with longer tenure are less likely to leave.
 Month-to-month contracts have the highest churn rate, while two-year
contracts show the least churn.
 Service Usage Insights:

 Customers with Fiber optic internet churn more than those with DSL or no
internet service. This may be due to service quality or pricing issues.
 Additional services like online security, tech support, and device
protection reduce churn, indicating that customers who invest in these
services are more likely to stay.
 Demographic Influence:

 Senior citizens have a higher churn rate than younger customers, possibly
due to price sensitivity or service adaptation challenges.
 Customers without dependents or partners are more likely to churn,
suggesting family or household stability plays a role in retention.
 Billing & Payment Trends:

 Electronic check payments correlate with higher churn, while credit card
and bank transfer payments see lower churn.
 Customers with higher monthly charges tend to leave more often,
highlighting potential pricing concerns.
 Correlation Analysis (Heatmap Insights):

 Tenure is negatively correlated with churn, reinforcing that longer-term


customers are more loyal.
 Contract type and payment method have strong associations with churn,
emphasizing the importance of structured contracts and
payment flexibility.

You might also like