Scribd
Upload
483 views20 pages

Customer Churn: by Dinesh Nair Adrien Le Doussal Fiona Tait Fatma Ahmadi Fulya Percin

This document discusses customer churn, including a case study on predicting customer churn in retail banking. The case study uses logistic regression on data from a Finnish bank to predict whether customers will churn. Key findings are that churn profiles depend on how a company values customers, and the model should be regularly updated to maintain prediction accuracy given the dynamic nature of customer behavior. Industries that commonly analyze churn include banking, telecommunications, and online retail.

Uploaded by

Fati Ahmadi
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
483 views20 pages

Customer Churn: by Dinesh Nair Adrien Le Doussal Fiona Tait Fatma Ahmadi Fulya Percin

This document discusses customer churn, including a case study on predicting customer churn in retail banking. The case study uses logistic regression on data from a Finnish bank to predict whether customers will churn. Key findings are that churn profiles depend on how a company values customers, and the model should be regularly updated to maintain prediction accuracy given the dynamic nature of customer behavior. Industries that commonly analyze churn include banking, telecommunications, and online retail.

Uploaded by

Fati Ahmadi
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 20

Customer Churn

by
Dinesh Nair
Adrien Le Doussal
Fiona Tait
Fatma Ahmadi
Fulya Percin
Definition
It’s History
Churn Types
Why it’s Important
It’s usage in today’s world
Industries that use it
 Banking sectors

 Telecommunication sectors
Retail Banking Case Study

Background
• The goal of this case study is to apply data mining techniques to
predict customer churn and analyze them by using data from a retail
banking company
• Once identified, these customers can be targeted with proactive
campaigns for retention efforts.

• Also and understanding the true value of a possible customer churn


will help the company in its CRM.
Retail Banking Case Study
Data Collection
 Different combinations of data hold different
analytical powers.

 Must indentify the data that best suits the type of analysis that is
being performed

 The quality of the data will determine the power and accuracy of the
overall model.
Retail Banking Case Study
 Which Data was collected ?
 Customer database from a Finnish bank is used and analyzed,
consisted only of personal customer.
 Time period from; December 2001-December 2005.
 The churn definition is based on current account
 In total 75 variables were collected and the related topics are as
follows; account transactions IN, account transactions OUT, service
indicators, personal profile info, customer level combined info.
 Had 251000 customers overall, and the data was divided into 3
groups randomly; samples are used for descriptive analysis,
regression analysis and potential future use in validation purposes.
Retail Banking Case Study
 Data Mining Technique; Logistic Regression
 Binary logistic regression is a form of regression which is used in a
situation when dependent is not a continuous variable
 To predict a discrete outcome on the basis of continuous and/or
categorical variables
 There is only one dependent variable. Dependent variable can take
value 1 with prob q and value 0 with prob 1-q.
 Results of the continuous probability are produced and divided into
2 groups(churners&nonchurners) by using threshold value which is
0.5.
Retail Banking Case Study
 Limitations
 Retail banking is characterized by many customers compared to
wholesale banking -This difference makes it hard to define customer
churn based on customer profitability
 This model reflect the dynamic nature of the churning customer
profile-The findings indicate that in the logistic regression, the user
should update the model to be able to produce predictions with high
accuracy
Retail Banking Case Study
 Results
The predicted churn customer profiles are presented to the bank but;

 Churn customer profiles depends on the company’s perspective


whether the churning customers are worth retaining or not.

 The customer churn analysis in this study might not be interesting if


the customers are valued based on the customer lifetime value
Data Collection methods
Data collection in Different
Industries
 Banking Industry
Transactions made by consumer
Products/services most likely to buy
Fraud detection
Customer most recent services and benefits desired
 Supermarket Industry
Identify unexpected shopping patterns
Defining new customer groups
Predict customer detections
Distinguish between profitable/non-profitable consumers
 Online Shopping
Consumer product purchase patterns
AMAZON= Buy Self-help book and Children book at the same time helps AMAZON detect this
as consumer buying pattern and may recommend childrens books to male aged 40 plus
who purchase Self Help books
Customer Churn Model
 Giving the right offer
 To the right person
 At the Right time
 Through the right channel

Source: Kurt Thearling


Collection Methods
 Logistic Regression is a type of regression that shows
whether or not a consumer will switch to an alternative
brand/service.

 It is binomial. Churn or no-churn

 Customer Loyalty= 1 - Customer Churn


Limitations..
Future Recommendations ..
Conclusions

You might also like