Professional Documents
Culture Documents
Group Report
SS 15
Bank Marketing
Vanessa Hofsttter
Patricia Mller
Konstantin Kremslehner
Sina Israel
Table of Content
Abstract
Direct Marketing is the practice of delivering promotional messages
directly to customers or prospects on an individual basis rather than using
a mass medium (Mandapaka et al., 2014). This paper suggests a binary
response model that helps a Portuguese financial service institute
targeting high quality receptive customers for their future campaigns and
identifies the important factors that influence the marketing campaigns
success to effectively manage resources. The research is based on a real
world sample data set from a Portuguese banking institute conducting a
direct telephone campaign (4,521 cases and 16 variables) containing past
campaign information (e.g. previous campaign outcome, duration of call or
customers personal and banking information). To predict the binary
customer responses to subscribe for a term deposit after being contacted
by a sales representative via telephone, a Logistic Regression Model was
built. The analysis showed that customers are more likely to subscribe for
a term deposit when they are well-educated, perform a rather risk-averse
job and are either single or divorced.
Introduction
Banks and other financial institutes operate in a highly competitive and
saturated market that is still lacking trust from both customer and inter
bank perspective after the global financial crisis in 2008. Especially the
Portuguese banking system was hit severely which contributed to the
European economic slowdown.
Due to a relative low interest rate for term deposits 1, banks are facing the
challenge to acquire customers who store their money for a fixed period of
time at the bank, which is a cost-efficient refinancing instrument for the
bank. Due to this present market condition, it is of utmost importance for
marketers to use a sophisticated Customer Relationship Management tool
1
A term deposit held at a financial institution has a fixed term. These are generally short-term
with maturities ranging anywhere from a month a few years. When a term deposit is purchased, the
lender (the customer) understands that the money can only be withdrawn after the term as ended
or by giving a predetermined number of days notice. (source: investopia.com)
Description
Job
Job description
Range of
Values
19-87
administrative
blue-collar
entrepreneur
housemaid
management
Type of
Information
Personal
Personal
retired
self employed
Marital Status
Education
Default
Balance
Housing
Loan
Contact
services
student
technican
unemployed
unknown
Marital Status of the
Divorced
consumer
Single
Married
Education of the consumer Primary
Secondary
Tertiary
Unknown
Indicates whether the
yes - no
client has a credit default
Account Balance
(-)3313 71188
Indicates whether the
yes - no
client has a housing loan
Indicates whether the
yes - no
client has a personal loan
Type of contact
cellular
communication
telephone
unknown
Day
Month
Duration
Campaign
pdays
Previous
Outcome
Jan - Dec
4 - 3025
1 - 50
(-) 1 - 871
0 - 25
Success
Personal
Personal
Bank
Bank
Bank
Bank
Information of
current campaign
Information of
current campaign
Information of
current campaign
Information of
current campaign
Information of
current campaign
Information of
current campaign
Information of
current campaign
Information of
previous
campaign
Information of
previous
campaign
Information of
previous
campaign
Failure
Other
Unkown
To reflect the frequencies within the data set, relevant predictor variables
will be analyzed:
Job
21,4 % of the data set (969 cases) are working in management positions
and thereby constitute the biggest group within the data set. Blue-collar
worker account for 20,9 % of the data (946 cases) and technicians with
17% (768 cases) are the third biggest group in the data set. In contrast,
students, housemaids and unemployed account for the lowest proportion
of the sample.
job
Frequenc
y
Percent
Valid admin.
478
10,6
blue-collar
946
20,9
entrepreneur
168
3,7
housemaid
112
2,5
management
969
21,4
retired
230
5,1
self183
4,0
employed
services
417
9,2
student
84
1,9
technician
768
17,0
unemployed
128
2,8
unknown
38
,8
Total
4521
100,0
Education
Valid
Cumulative
Percent
Percent
10,6
10,6
20,9
31,5
3,7
35,2
2,5
37,7
21,4
59,1
5,1
64,2
4,0
68,3
9,2
1,9
17,0
2,8
,8
100,0
77,5
79,3
96,3
99,2
100,0
Regarding the level of education, it can be seen that more than half of the
data set consists of customers with a secondary education, followed by
tertiary (29,9 %) and primary education (15 %)
education
Frequenc
y
Percent
Valid
Percent
Cumulative
Percent
Valid primary
secondar
y
tertiary
unknown
Total
678
15,0
15,0
15,0
2306
51,0
51,0
66,0
1350
187
4521
29,9
4,1
100,0
29,9
4,1
100,0
95,9
100,0
Marital Status
Married customers account for the biggest proportion in the data set (61,9
%), followed by single (26,5 %) and divorced (11,7 %).
marital
Frequenc
y
Percent
Valid divorced
528
11,7
married
2797
61,9
single
1196
26,5
Total
4521
100,0
Valid
Cumulative
Percent
Percent
11,7
11,7
61,9
73,5
26,5
100,0
100,0
housing
Valid no
yes
Total
Frequenc
y
Percent
1962
43,4
2559
56,6
4521
100,0
Valid
Cumulative
Percent
Percent
43,4
43,4
56,6
100,0
100,0
loan
Valid no
yes
Total
Frequenc
y
Percent
3830
84,7
691
15,3
4521
100,0
Valid
Cumulative
Percent
Percent
84,7
84,7
15,3
100,0
100,0
Sources
Lilien, Gary L., Philip Kotler, and K. Sridhar. Moorthy. Marketing Models.
Englewood Cliffs, NJ: Prentice-Hall, 1992.
Mandapaka, A., Kushwah, A. Role of Customer Response Models in
Customer Solicitation Centers Direct Marketing Campaign, Chakraborty
Oklahoma State University, OK, USA, 2014
Sulock, M. Application of Binary Logistic Regression to College Admission
Data, 2009