CHAPTER 1

INTRODUCTION

Chapter – 1
Introduction
_____________________________________
1.1 Problem Definition
Bank Marketing has become a necessary survival weapon and is fundamentally
dynamical in the banking system worldwide. The increase of Bank Marketing is
redefining business relationships and therefore the most made banks are those who will
really strengthen their relationship with their customers. Technology innovation and
fierce competition among existing banks have alter a large array of banking
services.Technology is neutering the relationships between banks and its internal and
external customers.

How will the financial organizations have a huge effectiveness for future promoting
campaigns? so as to answer this, we've to research the last Marketing campaign the bank
performed and establish the patterns that may facilitate conclusions so as to develop
future methods.

1.2 Motivation
Banking system is actually a industry that provides varieties of banking and allied
services to its clients. Bank customers are such persons and organizations that have
surplus or shortage of funds and those who need various types of financial and related
services provided by the banking sector. These customers belong to completely different
state of economy, geographical locations and different professions and businesses.
Naturally, the necessity of every individual cluster of clients is distinct from the
requirements of different teams. It is,therefore, necessary to spot completely different
solid teams and even sub-groups of clients, and so with utmost exactness verify their
desires, design schemes to suit their precise desires, and deliver the foremost
expeditiously.

The Role of selling within the banking system continues to alter. For several years the
first focus of bank promoting was publicity. Then the main target shifted to Advertising
and Marketing. That was followed by target the event of a sales culture. Though all the
weather of the promoting conception – client satisfaction, profit integrated framework
 and social responsibility – can stay necessary, client satisfaction should receive the best
stress within the years ahead. The chief considerations of most bank executives still target
legal and regulative problems, consistent with most surveys.

1.3 Background
Principal aspects of Bank Marketing

1.3.1 Customer Oriented Services

Services offered by the banks area unit to be found out in such a way that they fulfill the
requirements of the clients. historically, bankers are acquainted with assume in terms of
what banks offers and not what customers needs. So as to design and deliver client
required services, the banks should learn to hunt data regarding the prevailing and
potential customers, and their perceived and latent desires on an everyday and systematic
basis.

1.3.2 Design & Delivery of Such Services

The word design implies that sensible promoting services have to be compelled to be
properly designed and made thus on suit a selected well-defined cluster of customers.
Moreover, such properly designed services should be properly listed. the standard of
delivery is to be ensured not solely through targeted packaging, however additionally
through correct client services offered at the bank.

1.3.3 Corporate Objectives of the Bank

The corporate objectives of the bank area unit to be found out at intervals.The company
objectives area unit of two types, Short Term and Long Term.
 The Short Term Objectives may be of the type:-

a)Increasing gain of the bank next year.

b)Widening client base by providing new services,

c)Increasing rate of growth of credit next year,etc.

The Long Term Objectives might be: -

a) To rise to favored position in 5 years,

b) To become the universal bank over the amount of next 3years,etc.

Once the company objectives area unit clearly whear out, varied schemes will be
designed to meet the requirements of the customers at intervals the framework of the
chosen company objectives. Further, the resources created offered for systematic
promoting efforts are forced by policies, vision and attitudes of the management.

1.3.4 Environmental & Other Constraints

Generally, the environmental constraints fall under four categories:Economic, Cultural,

Legal and Political.

1.3.5 Marketing ideas – Its Application to Banking

When we apply marketing to the banking system, the bank marketing strategy will be a
fore mentioned to incorporate the following:

 A very clear definition of target customers.

 The Development of selling combine to satisfy customers at a profit for the bank.

 Planning for every of the supply markets.

 Organization and Administration.

1.3.6 Marketing Strategy in Banking Sector

 Consumer Behavior and Segmentation

Banks deal with individuals or group of people etc , all of whom have their likes and
dislikes. No bank can afford to assess the needs of each and every individual buyer
(actual or potential). Segmentation of the market into more or less homogenous groups,
in terms of their needs and expectations from the banking industry, provides a solution to
this problem.
 Identification of Target Customers & their desires
Unless the bank has clear idea about which type of customers it wants to serve,it is not
possible to work out products to satisfy their needs. This identification process might
involve

 Finding out the information such as the present age,sex,educational background,locality,

puchase habits etc of the present available customers.
 Taking a customer feedback for every few months, regarding the services provided.
 Collecting information from people who are not the actual customers of the bank.
All this may be done by conducting a survey of customers and non-customers of the
bank. Moreover, this method of seeking data regarding the market should kind associate
degree integral a part of the system and should be done on an everyday basis. The survey
would provide valuable data regarding profiles and opinions of customers and non-
customers of the bank,and it can be analyzed to find out the target group of the customers
and their felt and latent needs.

1.4 Existing System

1.4.1 Paid advertising

This includes multiple approaches for marketing. It includes traditional approaches like
TVCs and print media advertising.
1.4.2 Cause promoting
Cause promoting links to the products or services of a company to a social cause .

1.4.3 Relationship marketing

This type of marketing is basically focused on customer building and Enhancing existing
relationships with customers .

1.4.4 Undercover marketing

This type of promoting strategy focuses on promoting the product whereas customers
stay unaware of the marketing strategy. it's additionally called concealment promoting.

1.4.5 Transactional marketing

In transactional marketing the retailers encourage customers to buy with shopping

coupons, discounts and huge events. It enhances the chances of sales and motivates the
target audience to buy the promoted products.

1.5 Limitations of Existing System

1.5.1 Hard to get Momentum

While taking the time to develop a promoting strategy will be key to propel a company's
growth over a amount of your time, there area unit disadvantages of implementing a
promoting strategy. For one, it will be difficult to get momentum. although you'll have
associate degree innovative plan and a capable team to execute, there is not any guarantee
of success.
1.5.2 Cost

Executing made promoting methods typically takes time, cash and resources and for a
little company particularly, that may be frustrating.

1.5.3 Expose Your Weaknesses

In addition, creating a marketing strategic plan can also identify your weaknesses as a
company and as a business owner. It can also lead to faulty marketing decisions if
properly, and creates unrealistic financial projections if the information isn't interpreted
correctly.

1.6 Proposed System

This system presents a method of targeting customers for various types of banking
schemes using the advanced data set, which is prepared by using various data-analytical
methods and also clearly proves the life time of the clientele.
 CHAPTER 2
LITERATURE SURVEY
Chapter – 2

Literature Survey

2.1 Introduction

In this section of the related work we describe the existing system, the limitation of the
existing system, the earlier version, current version, proposed version, and expected
results of this project.

2.2 Related Paper Discussions

2.2.1 Title: A DATA MINING APPROACH TO PREDICT PROSPECTIVE BUSINESS

SECTORS FOR LENDING IN RETAIL BANKING USING DECISION TREE
Year: 2017
Author: Md. Rafiqul Islam and Md. Ahsan Habib2
Description:

This paper discusses about the implemented the procedure of a data mining task for
prospective business sectors analysis in retail banking. And also predefined the optimal
number of clusters (Excellent,Very Good,Good, Marginal and Bad)and used
classification methods(Decision Tree)to predict prospective business sectors based on
existing customer transactional behavioural data. The initial findings show some
interesting results. The field test conducted and found that the data mining prediction
model for prospective business sectors for disbursement of loan is very accurate and the
target-oriented campaign is very effective. The results show that most prospective
customers are in Rice and Flower Mils (Rice And FlowerMils) sectors; and Retail
Traders (RetailTraders) and Whole Seller (WholeSeller) are also very good.

The main focus of the paper was on written attributes of business customer but customer
have many unwritten behaviour attributes. So the topic exclude some important unwritten
 attributes such characters, conditions and collaterals. The model is developed based on
datamining method of decision tree classification as an initial step.

2.2.2 Title:The value of Data in Defining Strategies and Marketing Management

Year: 2018

Author: Jonas Schröder

Description:

The objective of this paper is to assess the value of data in defining marketing strategies
and marketing management. The technological advances in recent years offer many
opportunities to marketing practitioners and researchers. However, facing abundance in
data supply makes marketing analytics often similar to finding the famous needle in a
haystack. The first section focuses on the challenges that come with Big Data. In section
2, the author analyzes the relevance of a selection of data types: search data, user-
generated content, social network data, location and sensory data, clickstream data and
web-browsing behavior, and CRM data. Section 3 uses the 4Ps of the Marketing Mix as a
framework. For Product and Place, the author presents real business cases (Netflix and
Galeria Kaufhof). The subsection for Price will critically assess the possibilityofprice
discrimination. Regarding Promotion, the author will provide examples for how
companies can use data indirectly through advertising networks to target customers. The
overall result is presented in a table that connects data types and the4Ps.
 CHAPTER 3
SYSTEM ANALYSIS
 Chapter – 3
System Analysis
________________________________________

3.1 Introduction

Software Requirements Specification: It’s a description of a particular software product,

program or set of programs that performs a set of function in target environment. The
SRS contains the details of process, functions of the product, user characteristics. The
non functional requirements if any are also specified. The remaining section of the SRS
specifies the functionality of these systems. Further the need of Interfaces is also
described in the next part of SRS. The criteria for the non-functional requirements, the
constraints on the system and assumptions and dependencies (if any) are also described in
the remaining sections. 3.2 Software Requirement Specification.

3.2 Software Requirement Specification.

A Software Requirements Specification (SRS) is a description of a particular

software product, program or set of programs that performs a set of functions in a target
environment (IEEE Std. 830-1993). A Software Requirements Specification (SRS) is a
complete description of the behavior of the system to be developed. It includes a set of
use cases that describe all the interactions the users will have with the software. Use cases
are also known as functional requirements. In addition to use cases, the SRS also contains
non-functional (or supplementary) requirements. Non-functional requirements are
requirements which impose constraints on the design or implementation (such as
performance engineering requirements, quality standards, or design constraints).
3.2.1 Requirements Engineering.

Hardware Requirements
 System : Dell Inspiron 3543
 Hard Disk : 500 GB.
 Ram : 4 GB.

Software Requirements
The software requirements document is the specification of the system. It should
include both a definition and a specification of requirements. It is a set of what the system
should do rather than how it should do it. The software requirements provide a basis for
creating the software requirements specification. It is useful in estimating cost, planning
team activities, performing tasks and tracking the teams.
 Operating system : Linux Ubuntu 18.04
 Coding Language : Python
 Tool Used : Jupyter Notebook

3.3 Architecture

Figure 3.1 Data set preparation process

3.4 Module Description

Bank Data Set Preparation

Five Stages of Bank Data Set Preparation

1. In the first stage the bank data is collected from various sourses and stored in the form of a
csv file.
2. In the second stage, data is processed.
3. In the third stage, the missing values and the improper values are detected and necessary
corrections are made .
4. In the third stage the data which is now cleaned will be processed in the form of a file and
is ready to use for analysis.
5. In the Fourth Stage, the analysed data is put into different blocks according to the required
attributes. Now,this data is further used to make plans for a strategic marketing schemes.
 CHAPTER 4
IMPLEMENTATION & RESULTS
Chapter – 4

IMPLEMENTATION & RESULTS

_________________________________

4. 1 Marketing Introduction:
“The process by which companies create value for customers and build strong customer
relationships in order to capture value from customers in return.” - Kotler and
Armstrong (2010).
Marketing campaigns are characterized by focusing on the customer needs and their
overall satisfaction. There are 4 most important variables that should be considered when
planning for a marketing campaign.
The 4 Ps:
1) Segment of the Population: To which segment of people will this marketing campaign
be useful for. This is the most important segment ,which would be the only possible way
to reach out to the customers.
2) Distributing the channel to reach out to the customer's place: Which segment of the
population should we address? Which mode of communication should we use to get our
message out? (Ex: Telephones, Radio, TV, Social Media Etc.)
3) Price: What would be the best price that we can offer to potential clients?
4) Promotional Strategy: What strategy are we going to implement in order to target the
potential clients. This should be made the last part of the marketing campaign analysis
since there has to be an in depth analysis made available from the previous marketing
campaign. In our to know the changes that has to be made in the strategy planning for the
future campaingns.
4.2. What is a Term Deposit?
A Term deposit is a deposit that a bank or a financial institution offers with a fixed rate of
interest in which your money will be returned back after a specific maturity time.

4.3 Attribute Descriptions

Input variables:
i. bank client data:
1 - age: numeric data
2 - job: type of job (categorical: 'admin.','blue-
collar','entrepreneur','housemaid','management','retired','self-
employed','services','student','technician','unemployed','unknown')
3 - marital: marital status (categorical: 'divorced','married','single','unknown'; note:
'divorced' means divorced or widowed)
4 - education: (categorical: primary, secondary, tertiary and unknown)
5 - default: has credit in default? (categorical: 'no','yes','unknown')
6 - housing: has housing loan? (categorical: 'no','yes','unknown')
7 - loan: has personal loan? (categorical: 'no','yes','unknown')
8 - balance: Balance of the individual.
ii. Related with the last contact of the current campaign:
8 - contact: contact communication type (categorical: 'cellular','telephone')
9 - month: last contact month of year (categorical: 'jan', 'feb', 'mar', ..., 'nov', 'dec')
10 - day: last contact day of the week (categorical: 'mon','tue','wed','thu','fri')
11 - duration: last contact duration, in seconds (numeric). Important note: this attribute
highly affects the output target (e.g., if duration=0 then y='no'). Yet, the duration is not
known before a call is performed. Also, after the end of the call y is obviously known.
Thus, this input should only be included for benchmark purposes and should be discarded
if the intention is to have a realistic predictive model.
iii. other attributes:
12 - campaign: number of contacts performed during this campaign and for this client
(numeric, includes last contact)
13 - pdays: number of days that passed by after the client was last contacted from a
previous campaign (numeric; 999 means client was not previously contacted)
14 - previous: number of contacts performed before this campaign and for this client
(numeric)
15 - poutcome: outcome of the previous marketing campaign (categorical:
'failure','nonexistent','success')
Output variable (desired target):
21 - y - has the client subscribed a term deposit? (binary: 'yes','no')
 Figure 4.1 Importing and Checking the collected data.

4.4 Summary on distribution of data based upon various attributes

available:
 Mean Age ≈ 41 years old. (Min = 18 years old and Max =95 years old.)
 The mean balance is 1,528. However, the Standard Deviation (std) is a high number so we can
understand through this that the balance is heavily distributed across the dataset.
 Duration is obtained after the call is made to the potential client so if the target client has
never received calls this feature might not that useful. The reason why duration is highly
correlated with opening a term deposit is because the more the bank talks to a target client the
higher the probability the target client will open a term deposit since a higher duration means a
higher interest (commitment) from the potential client.
Figure 4.2 Checking attributes available in the loaded data

Figure 4.3 Checking for any missing values

 Figure 4.4 Code for information setup.

Figure 4.5 Result of the Information on Term Subscriptions

 Figure 4.6 Checking how attributes are distributed in the loaded data.

Figure 4.6.1 Checking how attributes are distributed in the loaded data and the deposits
information
Figure 4.7 Plot Diagrams for Amount of Type of Work and Balance by Term Subscription

Figure 4.7.1 Plot Diagrams for Distribution of Balance by Education

Figure 4.8 Loading the data in terms of the attribute “occupation”

4.5 Analysis by Occupation:
 Number : Management is the occupation that is more prevalent in this dataset.
 Age : As expected, the retired are the ones who have the highest median age while student are
the lowest.
 Balance : Management and Retirees are the ones who have the highest balance in their
accounts.

Figure 4.8.1. Distribution of data based upon the Term Job occupation
Figure 4.8.2. Mean balance in account by Job occupation
 Figure 4.8.3. Chart showing Number of Occupations

Figure 4.8.4. Candle stick graph showing Distribution of Ages by Occupation

 Figure 4.8.4(a)

Figure 4.8.4(b)

Figure 4.8.4 (a&b) Candle stick graph showing Distribution of Ages by Occupation in detail
with values.
4.6 Analysis by Marital Status:
As discussed previously, the impact of a divorce has a significant impact on the balance
of the individual.

Figure 4.9 Getting Analysis by Marital Status.

Figure 4.9.1 Bar Graph Showing Count by Marital Status

4.7 Analysis by Education:
The level of education also has a significant impact on the amount of balance a prospect
has.

Figure 4.10 Analysing data based upon Educational and Marital Grop
Figure 4.10.1 Horizontal Bar Graph showing Analysis of data based upon Educational and
Marital Grop
4.8 Analysis by Loans:
Whether the prospect has a previous loan has a significant impact on the amount of balance he
or she has.

Figure 4.10 Program for Analysis of data by Loans Data and related plot graph
Figure 4.10.1 Plot Graph showing impact of loans for various attributes.
Figure 4.10.1 Resultant Correlation Matrix to determine if duration has influence on term
deposits
 CHAPTER 5
TESTING AND FINAL ANALYSIS
 Chapter – 5

Testing and Final Analysis

_______________________________________

5.1 Introduction

Testing is the process of trying to discover fault or weakness in a developed work

product. It is a way to check the functionality of components, sub assemblies, assemblies
and/or a finished product.It is a process of achieving a desired work product in it’s
complete form without any drawbacks and to deliver a great working product.

In this project we are going to test the prepared analysed data components by combining
all the available attributes.

The following are the steps followed to do the final anlaysis

I) Classification Model

II) Confusion Matrix

III)ROC Curve

IV) Precision and Recall

5.2 Classification Model:

5.2.1 Stratified Sampling

It is an important concept that is often missed while developing a model either for
regression or a classification. In order to avoid over fitting of our data we must
implement a cross validation.We must atleast make sure that features which have the
mostof it’s influence on our label (whether a potential client will open a term deposit or
not) is equally distributed.
Personal Loans:
For example, having a personal loan is an important feature that determines whether a
potential client will open a term deposit or not. To confirm it has a heavy weight on the
final output we check the correlation matrix above and we can see it has a -11%
correlation with opening a deposit. What steps we should take before implementing
stratified sampling in our train and test data?
1) We need to see how our data is distributed.
2) After noticing that the column of loan contains 87% of "no" (Does not have personal
loans) and 13% of "yes" (Have personal loans.)
3) We want to make sure that our training and test set contains the same ratio of 87%
"no" and 13% "yes".

How can we avoid Overfitting?

The best availalbe alternative to avoid overfitting is to use cross validation. Taking the
training test data and splitting it. For example, if we split it by 3, 2/3 of the data or 66%
will be used for training and 1/3 or 33% will be used or testing and we will do the testing
process three times. This algorithm will iterate through all the training and test sets and
the main purpose of this is to grab the overall pattern of the data.
# Use Cross-validation.
from sklearn.model_selection import cross_val_score

# Logistic Regression
log_reg = LogisticRegression()
log_scores = cross_val_score(log_reg, X_train, y_train, cv=3)
log_reg_mean = log_scores.mean()

# SVC
svc_clf = SVC()
svc_scores = cross_val_score(svc_clf, X_train, y_train, cv=3)
svc_mean = svc_scores.mean()

# KNearestNeighbors
knn_clf = KNeighborsClassifier()
knn_scores = cross_val_score(knn_clf, X_train, y_train, cv=3)
knn_mean = knn_scores.mean()

# Decision Tree
tree_clf = tree.DecisionTreeClassifier()
tree_scores = cross_val_score(tree_clf, X_train, y_train, cv=3)
tree_mean = tree_scores.mean()

# Gradient Boosting Classifier

grad_clf = GradientBoostingClassifier()
grad_scores = cross_val_score(grad_clf, X_train, y_train, cv=3)
grad_mean = grad_scores.mean()

# Random Forest Classifier

rand_clf = RandomForestClassifier(n_estimators=18)
rand_scores = cross_val_score(rand_clf, X_train, y_train, cv=3)
rand_mean = rand_scores.mean()

# NeuralNet Classifier
neural_clf = MLPClassifier(alpha=1)
neural_scores = cross_val_score(neural_clf, X_train, y_train, cv=3)
neural_mean = neural_scores.mean()

# Naives Bayes
nav_clf = GaussianNB()
nav_scores = cross_val_score(nav_clf, X_train, y_train, cv=3)
nav_mean = neural_scores.mean()

# Create a Dataframe with the results.

d = {'Classifiers': ['Logistic Reg.', 'SVC', 'KNN', 'Dec Tree', 'Grad B CLF', 'Rand FC', 'Neural
Classifier', 'Naives Bayes'],
'Crossval Mean Scores': [log_reg_mean, svc_mean, knn_mean, tree_mean, grad_mean,
rand_mean, neural_mean, nav_mean]
result_df = pd.DataFrame(data=d)
 Figure 5.1 Resultant output of the classifiers and their Crossval Mean Scores

5.3 Confusion Matrix:

Figure 5.2 Confusion Matrix

Insights of a Confusion Matrix:

The primary purpose of a confusion matrix is to know how our model is performing
when it comes to classifying potential clients who are likely to subscribe to a term
deposit. We will see in the confusion matrix four terms the True Positives, False
Positives, True Negatives and False Negatives.

Positive/Negative: Type of Class (label) ["No", "Yes"] True/False: Correctly or

Incorrectly classified by the model.

True Negatives (Top-Left Square): This is the number of correctly classifications of

the "No" class or potential clients that are not willing to subscribe a term deposit.

False Negatives (Top-Right Square): This is the number of incorrectly classifications

of the "No" class or potential clients that are not willing to subscribe a term deposit.
 False Positives (Bottom-Left Square): This is the number of incorrectly classifications
of the "Yes" class or potential clients that are willing to subscribe a term deposit.

True Positives (Bottom-Right Square): This is the number of correctly classifications

of the "Yes" class or potential clients that are willing to subscribe a term deposit.

Figure 5.3
Related code and Resultant output of the Cross validation of the Gradient Boosting Classifier
 Figure 5.4 Resultant output in the form of a confusion matrix of the Cross validation of the
Gradient Boosting Classifier

5.4 Precision and Recall:

Recall: It is the total number of "Yes" in the label column of the dataset.

Precision: It tells us how sure is the prediction of our model that the actual label is a
"Yes".

Recall Precision Tradeoff:

As the precision gets higher the recall gets lower and vice versa. For example, if we
increase the precision from 30% to 60% the model is picking the predictions that the
model believes is 60% sure. If there is an instance where the model believes that is 58%
likely to be a potential client that will subscribe to a term deposit then the model will
classify it as a "No." However, that instance was actually a "Yes" (potential client did
suscribe to a term deposit.) That is why the higher the precision the more likely the model
is to miss instances that are actually a "Yes"!
 Figure 5.5 Related code and Resultant output of the Scores for precision and recall

5.5 ROC Curve (Receiver Operating Characteristic):

The ROC curve tells us how well our classifier is classifying between term deposit
suscriptions (True Positives) and non-term deposit suscriptions. The X-axis is
represented by False positive rates (Specificity) and the Y-axis is represented by the True
Positive Rate (Sensitivity.) As the line moves the threshold of the classification changes
giving us different values. The closer is the line to our top left corner the better is our
model separating both classes.
Figure 5.6 Related code to decide which threshold to use and return precision.
Figure 5.7 Resultant Graph of Precision and Recall Tradeoff
Figure 5.8 Related code of Gradient Boosting Classifier Curve.
Figure 5.9 Resultant Graph of ROC Curve Gradient Boosting Classifier.
Figure 5.10 Related code of Gradient Boosting Classifier,Neural Classifier,Naives Bayes
Classifier and their Scores.

Figure 5.11 Resultant Graph of ROC Curve of top 3 Classifiers.

 Figure 5.12 Prediction of the Gradient Boosting Classifier.

Which Features Influence the Result of a Term Deposit Suscription?

5.5 Decision Tree Classifier:

The top three most important features for our classifier are

 Duration (how long did the conversation between the sales representative and the
potential client)

 contact (number of contacts to the potential client within the same marketing campaign)

 month (the month of the year).

 Figure 5.5 Prediction of the Gradient Boosting Classifier.

Gradient Boosting classifier is the best model to predict whether or not a potential client will suscribe to
a term deposit or not. 83% accuracy!
CHAPTER 7
CONCLUSION
CHAPTER 7

CONCLUSION

1) Months of Marketing Activity: According toanalysis which we made, We saw that

the the month of MAY have the highest level of marketing activity.However, it was also
the same month in which the potential clientele rejected most of the offers for the
deposits,the Lowest Effective rate was about -34.49%. In order to overcome this For the
next marketing campaign, it is adviced for the bank to focus the marketing campaign
during the months of March, September, October and December. Considering
December to be ideal for effective marketing,due to less marketing rate.

2) Marketing Calls: A policy should be made,in which it states that no more tha calls
should be made to a client in oreder to save time and also to target different set of people.

3) Age Category: The bank marketing team should target people depending upon their
age as well. They should primarily target people belonging to early 20’s and also people
who are in their late 60’s. People under 20 and in early 20’s had a 60% chance of
suscribing to a term deposit while people belonging to their late 60’s had a 76% chance
of suscribing to a term deposit.

4) Occupation: As discussed in the above point people in early 20’s and old age people
in their 60’s who mostly are from student background and retired personell respectively
tend to suscribe more to the term deposit. Retired people tend to subscribe to this term
deposit in order to save their retirement amount and get a profitable intrest on that
amount.

5) House Loans and Balances: Potential clients with no balances and low balance tend
to have more number of loans such as house loans,personal loans etc because they are in
a phase where they could afford only to pay back the loan amount in installment bases
rather than saving them for term deposists. However, we can see that potential clients in
the average and high balances are less likely to have a house loan and therefore, more
likely to open a term deposit. Hence, the next marketing campaign should target
individuals of average and high balances in order to increase the likelihood of suscribing
to a term deposit.

6) Develop a Questionaire during the Calls: Since duration of the call is one of the
effective marketing strategy that most positively correlates with whether a client will save
their money in term deposit or not,by providing an intresting questionare to the clients
while making marketing calls may help us by getting some insights on the wants of a
customer.How ever, this does not assure us that the client will subscribe to a term
deposit! But, we don't loose anything by implementing a strategy that will increase the
level of engagement of the potential client leading to an increase probability of suscribing
to a term deposit, and therefore an increase in effectiveness for the next marketing
campaign the bank will execute.