Integrating multi-criteria decision making and clustering for business customer segmentation
| Date | 13 July 2015 |
| Published date | 13 July 2015 |
| Pages | 1022-1040 |
| DOI | https://doi.org/10.1108/IMDS-01-2015-0027 |
| Author | Hülya Güçdemir,Hasan Selim |
| Subject Matter | Information & knowledge management,Information systems,Data management systems |
Integrating multi-criteria decision
making and clustering for
business customer segmentation
Hülya Güçdemir
Department of Industrial Engineering,
Celal Bayar University, Manisa, Turkey, and
Hasan Selim
Department of Industrial Engineering, Dokuz Eylul University, Izmir, Turkey
Abstract
Purpose –The purpose of this paper is to develop a systematic approach for business customer
segmentation.
Design/methodology/approach –This study proposes an approach for business customer
segmentation that integrates clustering and multi-criteria decision making (MCDM). First, proper
segmentation variables are identified and then customers are grouped by using hierarchical and
partitional clustering algorithms. The approach extended the recency-frequency-monetary (RFM)
model by proposing five novel segmentation variables for business markets. To confirm the viability
of the proposed approach, a real-world application is presented. Three agglomerative hierarchical
clustering algorithms namely “Ward’s method,”“single linkage”and “complete linkage,”and a
partitional clustering algorithm, “k-means,”are used in segmentation. In the implementation, fuzzy
analytic hierarchy process is employed to determine the importance of the segments.
Findings –Business customers of an international original equipment manufacturer (OEM) are
segmented in the application. In this regard, 317 business customers of the OEM are segmented as
“best,”“valuable,”“average,”“potential valuable”and “potential invaluable”according to the cluster
ranks obtained in this study. The results of the application reveal that the proposed approach can
effectively be used in practice for business customer segmentation.
Research limitations/implications –The success of the proposed approach relies on the availability
andqualityofcustomers’data. Therefore, design ofan extensive customerdatabase managementsystem
is the foundation for any successful customer relationship management (CRM) solution offered by the
proposed approach. Such a database management system may entail a noteworthy level of investment.
Practical implications –The results of the application reveal that the proposed approach can
effectively be used in practice for business customer segmentation. By making customer segmentation
decisions, the proposed approach can provides firms a basis for the development of effective loyalty
programs and design of customized strategies for their customers.
Social implications –The proposed segmentation approach may contribute firms to gaining
sustainable competitive advantage in the market by increasing the effectiveness of CRM strategies.
Originality/value –This study proposes an integrated approach for business customer segmentation.
The proposedapproach differentiatesitself from its counterpartsby combining MCDM and clustering in
businesscustomer segmentation. In addition,it extends the traditional RFM modelby including five novel
segmentation variables for business markets.
Keywords Fuzzy AHP, Multi-criteria decision making, Business customer segmentation,
Data clustering
Paper type Case study
1. Introduction
Customer segmentation can be defined as division of a customer base into distinct
and internally consistent groups with similar characteristics. It allows companies
to develop different marketing strategies according to customer characteristics.
Industrial Management & Data
Systems
Vol. 115 No. 6, 2015
pp. 1022-1040
©Emerald Group Publishing Limited
0263-5577
DOI 10.1108/IMDS-01-2015-0027
Received 26 January 2015
Revised 7 April 2015
19 April 2015
Accepted 21 April 2015
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/0263-5577.htm
1022
IMDS
115,6
To execute customer segmentation, various techniques and variables are proposed.
Clustering, classification, self-organizing maps (SOM), evolutionary algorithms,
interaction detection methods and artificial neural network techniques are some
of the extensively used segmentation techniques. Among them, data clustering is the
most commonly used technique and the components of the recency-frequency-
monetary (RFM) model are the most commonly used variables for the customer
segmentation (Punj and Stewart, 1983). Data clustering can be defined as the
unsupervised classification of observations and it is based on grouping similar
observations in the same cluster. RFM model is the numeric expression of customer
behaviors and very effective for determining key customers. However, definition and
computation of these variables can change depending on the problem (Miglautsch,
2000). For instance, these variables can be defined as: recency, period since the last
purchase; frequency, number of purchases made within a certain period; and moneta ry,
total money spent during a certain period.
Many researchers employed RFM model in their segmentation studies. Among
them, Chan (2008) segments the customers of an automobile retailer by using genetic
algorithm (GA). Chiu et al. (2009) proposed a decision support system for market
segmentation that integrates conventional statistical analysis and two intelligent
clustering methods; SOM and particle swarm optimization (PSO). Cheng and Chen
(2009) joined the quantitative value of RFM variables and k-means clustering algorithm
into rough set theory. Dhandayudam and Krishnamurthi (2012) suggested a clustering
algorithm to overcome the difficulties of traditional clustering algorithms, and
segmented the customers of a fertilizer manufacturing company. Wei et al. (2013)
combined SOM and k-means clustering to segment customers of a hair salon in Taiwan.
Deng (2013) proposed an algorithm which is based on k-means clustering, PSO and
artificial bee colony to classify customers in e-commerce environment. The reader may
refer to Wei et al. (2010) for the applications of RFM model.
There exist some other studies extending the RFM model by including additional
variables. For instance, Li et al. (2011) added “relationship length”variable to
traditional RFM model and segmented the customers of a textile manufacturing firm
using a two-step clustering method; Ward with k-means. They defined the relationship
length as the time between the last and first transaction. However, this definition
does not give any information about the repetitiveness of the transactions of a
particular customer. Therefore, strength of the relationship should be considered in
addition to the relationship length. In addition, Hosseini et al. (2010) considered
“loyalty”besides the RFM variables. They divided the database into equal quintiles.
However, they do not apply any normalization method in their study. In addition, the
details on how to obtain the loyalty value were not given. Furthermore, the researchers
divided the customer base into 34 clusters, which is impracticable in most cases.
In some segmentation studies, problem-specific variables have been used instead
of RFM. For instance, Kim et al. (2006) displayed the customers of a wireless
telecommunication company with 3D space with axes denoting current value, potential
value and customer loyalty and segmented the customers. Teichert et al. (2008)
applied latent class modeling and segmented airline passengers based on behavioral
and socio-demographic variables. Ahn and Sohn (2009) identified customer groups
and propose suitable after sales services to these groups. They grouped customers by
using fuzzy c-means clustering along with the indicators of customer satisfaction index.
Gilboa (2009) segmented Israeli mall customers by using Ward with k-means.
In another study, a soft clustering method that uses a latent mixed class membership
1023
Integrating
MCDM and
clustering
To continue reading
Request your trial