BUILDING COMPETITIVE ADVANTAGE VIA CRM BASED ON
DATA WAREHOUSE AND DATA MINING
Jiejun Huang, Wei Cui, Yanbin Yuan
School of Resource and Environment Engineering, Wuhan University of Technology, Wuhan, China
Keywords: Customer Relationship Management (CRM), Data Mining, Data Warehouse, Decision-making.
Abstract: Customer Relationship Management (CRM) is providing a novel approach for managing the relationships
between a corporation and its customers towards maximal profitability and sustainability. Data mining and
data warehouse are the useful information technologies, which provide powerful means for extracting and
utilizing the business information from historical data resources and runtime data flows. This paper reviews
the objectives, functionalities, and development trends of CRM, discusses the architecture, data model and
development methodologies of CRM systems based on data warehouse and data mining, then outlines the
applications of integrated CRM systems in decision making, including business administration, marketing,
customer service, customer management, and credit evaluation. Eventually, it describes some problems and
challenges for further research.
1 INTRODUCTION
How to manage the enormous enterprise and the
involved customer group becomes a crucial problem
for corporation. It is a gigantic impact for global
economy, such as insolvency of Enron, financial
fraud of Worldcom, and the degenerate of Polaroid
and Xerox. For this reason, it is widely concerned
for all the enterprise to improve the corporation
management, and customer relationship
management. Data warehouse and data mining are
both promising new technologies to transact
information. It is their individual superiority to
organize the business macro-database, and to mine
the hidden and meaningful information. And it is
very helpful for increasing a company’s
effectiveness and enhancing the competitiveness
(Adam R, 2001).
The paper is organized as follows. Section 2
provides a brief summary of CRM systems. Section
3 discusses the structure and the process of CRM
system based on data warehouse and data mining. In
section 4, the commercial application of the
integration CRM system is deeply analyzed,
including business administration, market selling,
customer management and credit evaluating. At last,
we describe a set of challenging problems in section
5, and conclude with a summary.
2 WHAT IS CRM
2.1 The Goal and Function of CRM
CRM is aimed for management efficiency, and used
new technologies to manage relationship between
corporation and its customers effectively. The main
functions of CRM include analyzing the customer’s
purchase interests, classifying the customers, and
seeking latent and worthy ones. Thereby, the
corporation can carry out personality service;
heighten customer’s satisfaction and loyalty. In
order to keep competitive advantage, the corporation
must adopt the strategy of customer centred, and
build CRM system (Alex B, Smith,K 2000). It will
decrease selling-cost and provide scientific
conference for company to draw up production
strategy and development scheme.
Through the CRM, the company can adjust the
portal to connect with the customer, manage market
resource effectively and build more value customer
relationship, perform customer segmentation,
ascertain the target market. It establishes an
integrated feedback system, customer can
comprehend the company more easy, and the
company can provide superior service for them. All
that may lead to the relationship more closely
between the company and its customers.
287
Huang J., Cui W. and Yuan Y. (2006).
BUILDING COMPETITIVE ADVANTAGE VIA CRM BASED ON DATA WAREHOUSE AND DATA MINING.
In Proceedings of the Eighth International Conference on Enterprise Information Systems - AIDSS, pages 287-290
DOI: 10.5220/0002439402870290
Copyright
c
SciTePress
2.2 Trends and Directions of CRM
CRM was expounded in the end of 20
th
century. Its
systemic capability is enlarged constantly, and the
application domains are more extensive. In the
future, the CRM systems will be connected with
network, and become integration, intelligent and
automation.
2.2.1 Web CRM
CRM is the core parts of current company. With
more and more company becoming members of the
World Wide Web, It is well known that the Internet
will become even more vital. The Internet has
exploited a widely market across space and time, it
will be a basic platform for the development of
enterprises. The internal sources can be shared by
Web CRM, and the business process is optimized.
No doubt that web CRM will be the future direction;
it can support all the process, such as web
management, web service and web marketing.
2.2.2 Intelligent CRM
CRM combines automatic selling technologies and
intelligent systems, such as automatic withdrawal
system, intelligent query system and public
information service system, to achieve business
intelligent perfectly. The fusion of communication
methods (phone, fax, and email), automatic
processing the customer relationship, and building
automatic prediction model, all of these can provide
an intelligent method to make decision. The goal of
CRM will be automatic and intelligent system; it can
do all the trade automatically, including customer
management, process the business information,
advertise and sell activities.
2.2.3 Integrated CRM
There takes great progress in the combination of
CRM and ERP, and leads to infallible tendency for
development of enterprise (Allan R, 2002).
Meanwhile, new technologies apply in CRM,
including data warehouse, data mining and OLAP,
which will strengthen the function of CRM. Data
warehouse can efficiently manage and analyze the
complex business information. OLAP can be used to
analyze the data from multidimensional, visual and
complex view (Nordine M, 2001). Data mining may
discover the meaningful relationship, pattern, or
model from the customer data, which can guide the
business activities, including customer
segmentation, customer retention, and customer
scoring, etc.
3 BUILDING CRM SYSTEM
BASED ON DATA WAREHOUSE
AND DATA MINING
Data warehouse can help the company do better for
its customer service, and then create immense profits
(Gabrielle G, 1999), (Lariviere Bart, Van Den,
2005). More important that data warehouse has
allowed the company to strengthen CRM core
capabilities and business partnerships (Lawyer J,
Chowdhury S, 2004). Data mining provides an
information technology to develop and utilize the
data; it is very helpful for making decision by
extracting regulations, patterns and models from
large databases. And using knowledge discovery
techniques is favourable to reaching a competitive
advantage with CRM (Gottgtroy M, 2003).
Therefore, the CRM system based on data
warehouse and data mining has significant business
value, it components jointly as data collecting and
integrating, data modelling and knowledge
discovery, business application and decision making.
The architecture of CRM system based on data
warehouse and data mining is showed as figure 1.
And the three steps to implement the CRM system
described as following.
Internal data
Customer data
Market data
Data collect, clean, transform and loa
d
CRM
Server
Data warehouse
Information
share
Processes
optimize
Metadata
Knowledge
Database
Report and
Query
OLAP
Data
Minin
g
Data
anal
y
sis
Business
plan
Customer
retention
Customer
se
g
mentation
Customer
service
Decision
Making
Figure 1: The architecture of CRM system based on dat
a
warehouse and data mining.
ICEIS 2006 - ARTIFICIAL INTELLIGENCE AND DECISION SUPPORT SYSTEMS
288
3.1 Surveying Market and
Collecting Data
Define a business subject is a first step to make
decision. We should collect the data about customers
and market, and survey the market around the goal.
There are lots of methods to capture the data,
generally, as questionnaire, customer register table,
statistics data, financial report, Web log files and
tough points. The data not only includes customer
information, internal data, competitive data and
market information, but also includes the detailed
data such as cost record spending on customers,
transaction record and the customer contributions. In
order to insure the reliability of the data, it must be
cleaned and transformed and integrated. It is an
indispensable step for the system, and it will affect
the quality and effect of decision making directly.
3.2 Building Model and Extracting
Knowledge
Building data warehouse of CRM, including design
data structure, optimize its functions, store and
manage the metadata, etc. In order to employ OLAP
and make it easy to query and analyze data, we
should organize the data efficiently (showed as
fig.2). Then we can choose the method and
algorithm exactly to build data mining model, and
extract the hidden and meaningful knowledge. After
understanding the model, we should evaluate and
refine it until we have produced a model that is
likely to be successful in discovering knowledge.
The process is a core of the system which extracting
the meaningful information by data mining model.
3.3 Result Evaluation and Business
Application
This phase is to explain and evaluate the results, and
then apply the knowledge to specific domain.
Building the CRM server is to implement the
optimization of process, exchange and sharing of
information, feedback and transaction of customer
data. The system has many functions such as
analyzing the return of investment, the response
rates, and the churn rates. Based on the result, the
manager can draw up business plan and seek the
target market. Furthermore, they can carry out the
marketing strategy and adjust their production. By
the evaluation model, we can execute the customer
recognition, customer segmentation and customer
retention, order tracking, and step-selling service,
etc.
4 APPLICATIONS OF THE
INTEGRATED CRM SYSTEMS
4.1 Business Administration
From the point of system, the company can make all
of the processes optimization as customer service,
product development and marketing. It supports
information share and business cooperation, and
maximize the functions including resource
utilization and production. The company can find
requirement of customers, and explore a new
business domain and then improve the abilities of
market adaptation. With the CRM system, the
company can grasp the customers and markets more
intelligently. It has embodied an obvious result in
finance and telecommunication.
Product dimension
Product ID
Product name
Product category
Product price
Product time
Product description
Address dimension
Address ID
Street
City
State
Countr
y
Order ID
Order product
Order account
Order type
r
r
im
ID
Customer_ID
Product_ID
Transaction_ID
Order_ID
Total production
Total sailing
Total revenue
Fact table
Customer ID
Customer name
Customer phone
Customer record
Customer Grade
Address ID
Customer dimension
Transaction ID
Transaction date
Transaction location
Transaction price
Transaction type
Transaction dimension
Order dimension
Time dimension
Time ID
Day
Month
Quarter
Yea
r
Figure 2: Snowflake schema of the data for CRM.
BUILDING COMPETITIVE ADVANTAGE VIA CRM BASED ON DATA WAREHOUSE AND DATA MINING
289
4.2 Customer Service and Customer
Management
CRM has built a bridge for enterprise and customers,
and becomes more efficient and rapid by data
mining. With the right care and service, customers
will be high-value, and their profitability will be
optimized. After mining the customer’s hobbies and
interests, managers can improve business process.
This will reduce the cost, and raise the benefit.
Based on the information of customers, data mining
can help segment the prospective customers, seek
high profitable ones. Furthermore, it can create a
solution to solve the problems, such as which
customers are mostly to churn? The marketing tools
merged with data mining have a mighty ability to
predict the market, and business managers can
implement the marketing campaign (Groth R, 1999).
Companies can evaluate the marketing message to
be delivered to customers and identify high-value
ones. With personalized services, the relationship
between company and customers is kept tightly.
4.3 Prediction and Marketing
Forecast and exploit the market is a crucial matter in
improving economic benefit. CRM system can
provide online service for customers, and the
enterprise may quickly process the feedback about
product function and selling service, and catch the
information about interaction record, transaction
history and request. After that, marketing and
serving automatically will be come true, so the
quality and efficiency of the actions has been
improved. Web mining is helpful for identifying the
profitable customers, improving the respond rate and
reducing the cost. We can plan and predict the
marketing campaigns after mining the customer
data, implement positive selling and cross-selling,
then provide the manifold and intelligent service. So
that the business process has been expand and the
quality and efficiency of the actions has been
improved.
4.4 Credit Evaluation and Fraud
Detection
Based on the transaction records, we can score the
customer’s credit rate, and then carry out selling and
preferential service, enhance the abilities of market
responding. In insurance, managers can analyze the
action and feature of customers through building the
evaluation model of customer. Then, they survey the
instance and detect fraud, to discover who claims for
compensation with perjury and predict a potentially
fraudulent transaction. In telecommunication, data
mining can predict and track cellular fraud for
telecommunications, this control and reduces the
unexpected risk effectively.
5 FUTURE RESEARCH
CRM is the core parts of current company, how to
manage the customer relationship becomes a key of
business competition. In order to take the lead of
economic society, the company should make full use
of the information technologies, and utilize its
vantages and features. Data warehouse and data
mining have mighty abilities to analyze the
information, this provide a strong science evidence
and technological support. The integrated CRM
system will play an important role in commercial
activity; and we believe that it will have a widely
development market and application prospect.
ACKNOWLEDGEMENTS
The work in this paper was supported by National
Natural Science Foundation of China (60175022,
40571128, and 40572166)
.
REFERENCES
Adam R. (2001). CRM Shifts to Data Mining to Keep
Customers, Global Finance, 15 (11), 97-98.
Alex B, Smith,K(2000). Building data mining applications
for CRM. McGraw-Hill Companies,Inc.
Allan R. (2002). Integrating ERP Can Overcome CRM
Limits J, Software Magazine, 22 (1), 29-32.
Gabrielle Gagnon. (1999). Data Warehousing: An
Overview, PC Magazine, 18 (5), 245-246.
Gottgtroy M, Gottgtroy P. (2003). Ontologies, CRM, data
mining: How to integrate? Management Information
Systems, 7, 307-316.
Groth R. (1999). Data Mining: Building competitive
advantage,Prentice-Hall Ptr.
Lariviere Bart, Van Den poel Dirk. (2005). Predicting
customer retention and profitability by using random
forests and regression forests techniques, Expert
Systems with Applications, 29(2): 472-484.
Lawyer J, Chowdhury S. (2004). Best practices in Data
Warehousing to support business initiatives and needs.
Proc. of the 37th Annual Hawaii International
Conference on System Sciences, 3499-3507.
Nordine Melab. (2001). Data mining: A key contribution
to E-business. Information & Communications
Technology Law, 10(3), 309-318.
Ron K, Foster P. (2001). Applications of Data Mining to
Electronic Commerce, Data Mining and Knowledge
Discovery, 5(1/2), 1-7.
ICEIS 2006 - ARTIFICIAL INTELLIGENCE AND DECISION SUPPORT SYSTEMS
290