Propensity Models

Once segmentation is done, the next step is to identify and target the ideal segment. Firms usually pursue
everyone in their target segments with all their resources available. This may not be effective. In this case, it is
best to know which customers are more likely to respond to your campaigns or purchase your product.
Propensity models are one such methods used to identify those type of customers.

Propensity models are used in marketing to predict the behaviour of the customers and leads. You also learnt
about two types of propensity models.

1. Propensity to buy model: This model is mainly used for the prospect customers or leads. This will help
the firms to predict the likelihood of a lead buying your product or services. This method is also called
lead scoring or lead qualification. This will help you determine and prioritise the leads and thereby
increasing your lead conversion rate.
2. Propensity to churn model: This model is mainly used for the existing customers. This will help the
firm to predict the likelihood of a customer leaving your service or not making a repeat purchase from
you. This will help you in developing targeting and retargeting strategies for these customers.

You also understood what churn analysis is and learnt the definition of churn in different industries. You then
came to know that the classification models are used in developing the propensity models. one of the most
commonly used classification models is the logistic regression model. In the next segment, you will learn more
about logistic regression.

Linear regression can be used to predict dependent variable which is continuous in nature. But, in real life,
there are many situations where you are required to predict a categorical variable such as taking a yes or no
decision. In this case, linear regression is not capable of segregating the numerical predictions into categorical
discrete values.

© Copyright. UpGrad Education Pvt. Ltd. All rights reserved

Logistic Regression: Logistic regression is a statistical method to predict or classify the dependent variable
which is categorical in nature based on one or more independent variables.

• Dependent variable(Y): The dependent variable in logistic regression is categorical in nature. For
example, if the problem is to take an yes or no decision, then the Dependent variable Y takes values 0 or 1
o Y =1, if the decision is yes.
o Y = 0, if the decision is No

• Independent Variable(X): In logistic regression, the independent variables can be continuous or/and
categorical variables.

Logistic Regression equation

The form of regression equation used in logistic regression is

Unlike linear regression, in logistic regression, the dependent variable is not predicted directly. Since the
dependent variable can only take values 1 or 0, The probability of Y=1 is calculated.

The values of 0, 1 and 2 are estimated from the sample data and used for further classification of
dependent variable.

Confusion Matrix:
Confusion matrix is one of the methods to evaluate the performance and accuracy of the logistic regression
model. This evaluation of the logistic regression model is done by using the sample data. This sample data
contains the information of independent and known dependent variable. This known dependent variable is
compared with the dependent variable estimated by logistic regression model.

• If the logistic regression model classifies the dependent variable Y=1, then it is considered as
‘Positive’​.
• If the logistic regression model classifies the dependent variable Y=0, then it is considered as
‘Negative’.
• If the prediction of the dependent variable is same as actual known dependent variable, Then the
prediction is said to be ​‘True’
• If the prediction of the dependent variable is different from actual known dependent variable, Then
the prediction is said to be ​‘False’

© Copyright. UpGrad Education Pvt. Ltd. All rights reserved

The metrics to measure the Performance of the logistic regression model

1. ​Accuracy​: It measures how often the model is correct

a. Accuracy =​ Total True Predictions​ = TP + TN

Total Predictions Total Predictions

2. ​Error rate or Misclassification rate​: It measure how often the model is wrong .

​Error rate or Misclassification rate= ​Total wrong predictions​ = ​FP + FN

Total Predictions Total Predictions

3. ​Specificity​: It measures how often the model correctly predicts when the actual classification is no .

Specificity = ​Number of actual ′Y=0′ correctly predicted​ = ​TN

Total number of actual ′Y=0′ TN + FP

4. ​Sensitivity​: It measures how often the model correctly predicts when the actual classification is yes

Sensitivity = ​ Number of actual ′ Y=1′ correctly predicted ​ = ​TP

Total number of actual ‘Y =1’ TP + FN

© Copyright. UpGrad Education Pvt. Ltd. All rights reserved

By going through this session

● Understand the different types of segmentation

● Perform RFM analysis and segment the customers
● Develop targeting and communication strategy
● Understand and interpret K-means and hierarchical clustering
● Analyse data using story telling framework

Disclaimer: All content and material on the UpGrad website is copyrighted material, either belonging to
UpGrad or its bonafide contributors and is purely for the dissemination of education. You are permitted to
access print and download extracts from this site purely for your own education only and on the following
basis:

● You can download this document from the website for self-use only.
● Any copies of this document, in part or full, saved to disc or to any other storage medium may only be
used for subsequent, self-viewing purposes or to print an individual extract or copy for
non-commercial personal use only.
● Any further dissemination, distribution, reproduction, copying of the content of the document herein
or the uploading thereof on other websites or use of content for any other commercial/unauthorized
purposes in any way which could infringe the intellectual property rights of UpGrad or its
contributors, is strictly prohibited.
● No graphics, images or photographs from any accompanying text in this document will be used
separately for unauthorised purposes.
● No material in this document will be modified, adapted or altered in any way.
● No part of this document or UpGrad content may be reproduced or stored in any other web site or
included in any public or private electronic retrieval system or service without UpGrad’s prior written
permission.
● Any rights not expressly granted in these terms are reserved.

© Copyright. UpGrad Education Pvt. Ltd. All rights reserved