Customer Relationship Management and Knowledge

902

Category: Business Information Systems

Customer Relationship Management and Knowledge Discovery in Database Jounghae Bang Kookmin University, Korea Nikhilesh Dholakia University of Rhode Island, USA Lutz Hamel University of Rhode Island, USA Seung-Kyoon Shin University of Rhode Island, USA

IntroductIon

Background

Customer relationships are increasingly central to business success (Kotler, 1997; Reichheld & Sasser, 1990). Acquiring new customers is five to seven times costlier than retaining existing customers (Kotler, 1997). Simply by reducing customer defections by 5%, a company can improve profits by 25% to 85% (Reichheld & Sasser, 1990). Relationship marketing—getting to know customers intimately by understanding their preferences—has emerged as a key business strategy for customer retention (Dyche, 2002). Internet and related technologies offer amazing possibilities for creating and sustaining ideal customer relationships (Goodhue, Wixom, & Watson, 2002; Ives, 1990; Moorman, Zaltman, & Deshpande, 1992). Internet is not only an important and convenient new channel for promotion, transactions, and business process coordination; it is also a source of customer data (Shaw, Subramaniam, Tan, & Welge, 2001). Huge customer data warehouses are being created using advanced database technologies (Fayyad, PiatetskyShapiro, & Smyth, 1996). Customer data warehouses by themselves offer no competitive advantages: insightful customer knowledge must be extracted from such data (Kim, Kim, & Lee, 2002). Valuable marketing insights about customer characteristics and their purchase patterns, however, are often hidden and untapped (Shaw et al., 2001). Data mining and knowledge discovery in databases (KDD) facilitate extraction of valuable knowledge from rapidly growing volumes of data (Mackinnon, 1999; Fayyad et al., 1996). This article provides a brief review of customer relationship issues. The article focuses on: (1) customer relationship management (CRM) technologies, (2) KDD techniques, and (3) Key CRM-KDD linkages in terms of relationship marketing. The article concludes with the observations about the state-of-the-art and future directions.

crm technologies CRM is interpreted in a variety of ways (Goodhue et al., 2002; Winer, 2001; Wright, 2002). In some cases, CRM simply entails direct e-mails or database marketing. In other cases, CRM refers to CICs (customer interaction centers) and OLAP (online analytical processing), which is referred to as various types of online query-driven analyses for examining stored data. Overall, CRM can be seen as a core business strategy to interact with, create, and deliver value to targeted customers to improve customer satisfaction and customer retention at a profit. It is grounded in high quality customer data and enabled by information technology (Ang & Buttle, 2002). Three core dimensions characterize buyer-focused CRM systems: customers, management, and technologies. Customer service and related issues must be included in the design, implementation, and operation of any CRM system. Organizations benefit from CRM only when such systems benefit their customers—using CRM merely as a sales or customer service solution is a recipe for failure (Davids, 1999). Management’s articulation and tracking of customer relationship goals, plans, and metrics is an essential CRM component (Ang & Buttle, 2002; Greenberg, 2002). Successful CRM implementations rely on management goals, strategies, and plans that reflect customer commitment and promote a customer-responsive corporate culture at all levels of the organization (Ang & Buttle, 2002; Smith, 2001). Technologies for facilitating collaborative, operational, and analytical CRM activities are the manifest aspects of CRM (Goodhue et al., 2002). Collaborative CRM systems refer to any CRM function that provides a point of interaction between the customer and the marketing channel (Greenberg, 2002). Web-based

Copyright © 2009, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.

Customer Relationship Management and Knowledge Discovery in Database

Figure 1. Alignment of three dimensions of CRM Technological Dimension

Management Goal Strategy

Collaborative Technology Analytical Technology

Customer

Plan Operational Technology

non, 1999). As a component of KDD (Fayyad et al., 1996), data mining can be defined as the process of searching and analyzing data in order to find latent but potentially valuable information (Shaw et al., 2001). KDD constitutes the overall process of extracting useful knowledge from databases. It is a multidisciplinary activity with the following stages (Brachman, Khabaza, Kloesgen, Piatetsky-Shapiro, & Simoudis, 1996; Bruha, Kralik, & Berka, 2000; Fayyad et al., 1996): • • •

Internet, and in some cases mobile commerce systems, offer multiple “touch points” for reaching the customers. In employing the Web and mobile technologies, it is important to ensure that such technologies enhance older, preexisting channels (Johnson, 2002). Operational CRM systems refer to technologies that span the ordering-delivery cycle (Goodhue et al., 2002). Operational CRM is concerned with automating the customer-facing parts of the enterprise (Ang & Buttle, 2002). Since the sales process depends on the cooperation of multiple departments performing different functions, integration of all such functions is critical for operational CRM systems (Earl, 2003; Greenberg, 2002). Analytical CRM systems analyze customer data warehouses so that the firm can detect valuable patterns of customers’ purchasing behavior. Off-line data mining of customer data warehouses as well as online analytical processing (OLAP) can aid in applications such as campaign management, churn analysis, propensity scoring, and customer profitability analysis (Goodhue et al., 2002). It is this component of CRM that has a clear linkage to KDD methods.

kdd techniques Since multiple data formats and distributed nature of knowledge on the Web make it a challenge to collect, discover, organize, and manage CRM-related customer data (Shaw et al., 2001), KDD methods are receiving attention in relationship marketing contexts (Fayyad et al., 1996; Mackinnon, 1999). Massive databases are commonplace, and they are ever growing, dynamic, and heterogeneous (Mackinnon & Glick, 1999). Systematic combining of data mining and knowledge management techniques can be the basis for advantageous customer relationships (Shaw et al., 2001). KDD is defined as the process of data selection, sampling, pre-processing, cleaning, transformation, dimension reduction, analysis, visualization, and evaluation (Mackin-

• •

Selecting the problem area and choosing a tool for representing the goal to be achieved Collecting the data and choosing tools for representing objects (observations) of the dataset Preprocessing of the data: integrating and cleaning data Data mining: extracting pieces of knowledge Post-processing of the knowledge derived: testing and verifying, interpreting, and applying the knowledge to the problem area at hand

In Web-based relationship marketing, three distinct categories of data mining have emerged: Web content mining, Web structure mining, and Web usage mining (Jackson, 2002). Web usage mining is also referred to as clickstream analysis (Edelstein, 2001). Valuable information hidden in the clickstream data of many e-commerce sites can provide sharp diagnostics and accurate forecasts, allowing e-commerce sites to profitably target and reach key customers (Moe & Fader, 2001). Therefore, many detailed studies have been conducted on Web usage mining. For example, Web access pattern tree (WAP-tree) mining is one of the sequential pattern mining techniques for Web log access sequences (Ezeife & Lu, 2005). Such Web-based CRM systems require large, integrated data repositories and advanced analytical capability. Even though there are many success stories, Web-based CRM projects continue to be expensive and risky undertakings. OLAP refers to the various types of query-driven analysis for analyzing stored data (Berry & Linoff, 1997). Data mining and OLAP can be seen as complementary tools (Jackson, 2002). Both Web-based CRM systems and OLAP, in general, involve vast volumes of both structured and unstructured data. One common challenge with managing this data is to incorporate unstructured data into a data warehouse. Traditional database systems are not designed for unstructured data. Research in KDD in general is intended to develop methods and techniques to process a large volume of unstructured data in order to retrieve valuable knowledge (which is “hidden” in these databases) that would be compact and abstract, yet understandable and useful for managerial applications (Bruha et al., 2000).

903

C


Figure 2. CRM and KDD process connection CR

Collaborative

CRM

Operational

CRM

Analytical Customer

CRM

C U S T O M E R S

KDD

Defining problem & Collecting Cleaning Analyzing

Data mining

Interpreting

Cluster analysis Neural networks Regression analysis Decision Trees Discriminant analysis Correlation analysis Association Rules, etc. Information Flow

strengtHenIng crm-kdd lInkages Figure 2 explains the CRM-KDD linkage from a process point of view. As explained previously, the importance of gaining knowledge has been well recognized. In line with this notion, CRM starts with understanding customers and gaining in-depth knowledge about customers. Therefore, the intersection between KDD and CRM can be seen as the analytical CRM part of CRM systems and customer knowledge discovery in database process of overall KDD process as shown in Figure 2. Collaborative CRM systems help collect accurate information from customers, while operational CRM can capitalize on the result of the analyses. The problem definition stage of KDD process can be done also in the management dimension of CRM. Following the definition of KDD and data mining, techniques for data mining are included under the analysis stage of KDD. Turning to the relationship marketing issues, with help of CRM and KDD technologies, database marketing and one-to-one marketing methods have come to the fore. Direct and database marketing methods can be regarded as powerful instruments to achieve CRM goals even though CRM is not a sub-task of direct and database marketers (Wehmeyer, 2005). The strategic goal of database marketing is to use collected information to identify customers and prospects as individuals and build continuing personalized relationships with them, leading to greater benefits for the individuals and greater profits for the corporation (Kahan, 1998). Database 904

marketing anticipates customer behavior over time and reacts to changes in the customer’s behavior. Database marketing identifies unique segments in the database reacting to specific stimuli such as promotions (McKim, 2002). One-to-one marketing represents the ultimate expression of target marketing—market segments with just one member each—or at least one at a time (Pitta, 1998). It relies on a two-way communication between a company and its customers to enhance a true relationship and allows customers to truly express the desires that the company can help fulfill (Dyche, 2002). A promising solution to implementing one-to-one marketing is the application of data mining techniques aided by information technology. Data mining allows organizations to find patterns within their internal customer data. Whatever patterns are uncovered can lead to target segmentations. Armed with such information, organizations can refine their targets and develop their technology to achieve true one-to-one marketing (Pitta, 1998). As an extension of one-to-one marketing, the concept of permission marketing is focused on seeking customers’ consent about desired marketing methods. Customers not only need to be communicated with as individuals, but they themselves should also be able to stipulate how and when they wish to be approached (Newell, 2003). One-toone and permission marketing rely heavily on information technology to track individual customers, understand their differences, and acknowledge their interaction preferences (Dyche, 2002).


Table 1. Customer relationship related data analysis and data mining tools

C

CUSTOMER RELATIONSHIP MARKETING ISSUES

Issue

Challenge

Possible analysis

Data mining technique most likely used

Database Marketing

One-to-One Marketing

Permission Marketing

Understanding customers with the database on customer behavior over time including reactions to changes

Communicating with customers as individuals Developing custom products and tailored messages based on customers’ unspoken needs

Seeking customers’ agreement about desired marketing methods

Identifies unique segments in the database

Find patterns within the internal customer data. Track individual customers Understand their differences

Track individual customers Understand their differences Acknowledge their interaction preferences Stimulate the customer’s response

Classification Prediction

Classification Dependency Analysis

Regression analysis Neural networks Decision Trees Discriminant Analysis

Descriptive and visualization Neural networks Regression analysis Correlation analysis Decision Trees Discrimination analysis Case-based reasoning Association Rules

Segmentation

Descriptive and visualization Cluster analysis Neural networks

Data mining methods allow marketers to sift through growing volumes of data and to understand their customers better. Shaw et al. (2001) introduced three major areas of application of data mining for knowledge-based marketing: (1) customer profiling, (2) deviation analysis, and (3) trend analysis. Also, Jackson (2002) noted that data mining can be used as a vehicle to increase profits by reducing costs and/or raising revenue. Some of the common ways to use data mining in customer relationship contexts include: • •

Eliminating expensive mailings to customers who are unlikely to respond to an offer during a marketing campaign. Facilitating one-to-one marketing and mass customization opportunities in CRM.

In sum, many organizations use data mining to help manage all phases of the customer lifecycle, and CRM systems can benefit from well-managed data analysis based on data mining. Table 1 summarizes the relationship marketing issues, and includes the possible customer analyses and relevant data mining techniques. Though data mining can be very useful in finding hidden patterns, there are important notions to keep in mind. First, it is important to note that KDD and data mining involve continuous human-computer interactions (Firestone,

2005). Data mining is not like a simple easy button. It is the skills and background knowledge of humans that will make a difference in the performance of data mining and KDD. For example, Firestone (2005) argued that the quality of information in the data warehouse is critical. Records and information management managers can enhance the performance of KDD process with the ability to create information that can lead to new knowledge by interpreting and evaluating. Second, a fit between information technology (IT) in an organization and organizational strategy and marketing requirement is required (Wehmeyer, 2005). Relationship marketing and CRM have been mainly discussed at a strategic level (Wehmeyer, 2005). Therefore, rather than separated marketing support technologies functioning only for itself, there should be a fit (Venkatraman, 1989) between IT support and marketing requirements at the operational as well as at the strategic level to design successful IT-enhanced marketing processes (Wehmeyer, 2005).

future trends Due to the advance of information technology, there are more opportunities to collect the data about customers. The Internet provides and promotes interactive communications 905


between the business and the customers, and it leads to increasing volume of rich data about customers. Based on the rich data collected, it is possible to successfully identify new prospective customers by using customer lifetime value to evaluate the current customers and match their profiles with those of new prospects (Wilson, 2006). However, in interactive marketing contexts, customers are also able to “block out” the intrusive marketing actions, and therefore, appropriate depth and width of “permissions” should be obtained (Godin, 1999; Krishnamurthy, 2001). Therefore, understanding customers will become more critical as new information technology is being developed. Furthermore, companies and customers can have opportunities to co-create products, pricing, and distributions. Information technology provides this opportunities by allowing companies to assess each customer individually and then to determine whether to serve that customer directly or via a third party, and whether to create an offering that customizes the product or standardizes the offering (Sheth, Sisodia, & Sharma, 2000). All the decisions to make should be based on thorough analyses of customer data and accurate knowledge generation about customers. In the similar vein, it is projected that knowledge generation—business intelligence—and data mining technologies are integrated because of importance of business performance management (BPM), which is focused on metrics development and data gathering to measure performance (Firestone, 2005). Not only knowledge generation, but also knowledge sharing and dissemination through the organization should be considered. Shaw et al. (2001) argued that ownership and access to the marketing knowledge, standards of knowledge interchange, and sharing of applications become critical. Ho and Chuang (2006) argued that it is important to establish a knowledge management and CRM mechanism that has a system, a plan, a classification feature, objectives, and an evaluation management mechanism to satisfy customers. In various organizational environments, both managing KDD processes to generate customer knowledge and managing customer relationship based on the knowledge generated and shared through organization are challenges for the future.

This article highlighted some of the intersections between the two. Different relationship marketing issues have emerged, and these rely increasingly on CRM and KDD technologies, especially for in-depth analysis. Various data mining techniques and KDD processes exist and provide the right tools to solve relationship marketing problems. While companies are eager to learn about their customers by using data mining technologies, it is very difficult to choose the most effective algorithms for the diverse range of problems and issues that marketers face (Kim et al., 2002). Even though Table 1 illustrates the main relationship marketing issues, challenges, and the potential analytic data mining tools, it is the analyst who decides creatively which tool is appropriate for what in which situation and how to interpret the results. From a process point of view as well, gaining customer knowledge becomes critical for managing customer relationships, and systematic knowledge generating processes are of great benefit. For effective customer-centric marketing strategies, the discovered knowledge has to be managed in a systematic manner.

conclusIons

Firestone, J. M. (2005). Mining for information gold. Information Management Journal, 39(5), 47.

This article offered a brief review of customer relationship issues. CRM systems consist of management, technology, and customer dimensions. CRM technologies are divided into three categories: analytical, operational, and collaborative CRM. As the importance of knowledge increases, KDD techniques are receiving attention as systematic processes to generate knowledge. Although CRM and KDD began separately, the two concepts have points of convergence. 906

references Berry, M. J. A., & Linoff, G. (1997). Data mining techniques: For marketing, sales, and customer support. New York: John Wiley & Sons. Brachman, R. J., Khabaza, T., Kloesgen, W., Piatetsky-Shapiro, G., & Simoudis, E. (1996). Mining business databases. Communications of the ACM, 39(11), 42-48. Davids, M. (1999). How to avoid the 10 biggest mistakes in CRM. The Journal of Business Strategy, 20(6), 22. Dyche, J. (2002). The CRM handbook: A business guide to customer relationship management. Boston: AddisonWesley. Fayyad, U., Piatetsky-Shapiro, G., & Smyth, P. (1996). The KDD process for extracting useful knowledge from volumes of data. Association for Computing Machinery. Communications of the ACM, 39(11), 27.

Goodhue, D. L., Wixom, B. H., & Watson, H. J. (2002). Realizing business benefits through CRM: Hitting the right target in the right way. MIS Quarterly Executive, 1(2), 79-94. Greenberg, P. (2002). CRM at the speed of light: Capturing and keeping customers in Internet real time (2nd ed.). Berkeley and London: McGraw-Hill.


Jackson, J. (2002). Data mining: A conceptual overview. Communications of the Association for Information Systems, 8(2002), 267-296.

Wilson, R. D. (2006). Developing new business strategies in B2B markets by combining CRM concepts and online databases. Competitiveness Review, 16(1), 38-45.

Johnson, L. K. (2002). New views on digital CRM. Sloan Management Review, 44(1), 10.

key terms

Kim, E., Kim, W., & Lee, Y. (2002). Combination of multiple classifiers for the customer’s purchase behavior prediction. Decision Support Systems, 34(2002), 167-175. Krishnamurthy, S. (2001). A comprehensive analysis of permission marketing. Journal of Computer-Mediated Communication, 6(2). Moe, W. W., & Fader, P. S. (2001). Uncovering patterns in cybershopping. California Management Review, 43(4), 106-117. Moorman, C., Zaltman, G., & Deshpande, R. (1992). Relationships between providers and users of market research: The dynamics of trust within and between organizations. Journal of marketing research, 24(August), 314-328. Newell, F. (2003). Why CRM doesn’t work: How to win by letting customer manage the relationship. NJ: Bloomberg Press. Shaw, M. J., Subramaniam, C., Tan, G. W., & Welge, M. E. (2001). Knowledge management and data mining for marketing. Decision Support Systems, 31(2002), 127-137. Sheth, J. N., Sisodia, R. S., & Sharma, A. (2000). The antecedents and consequences of customer-centric marketing. Journal of Academy of Marketing Science, 28(1), 55.

Clickstream Data: Web usage data. A virtual trail that a user leaves behind while surfing the Internet. For example, every Web site and every page of every Web site that the user visits, how long the user was on a page or site. Customer Relationship Management (CRM): A core business strategy that promotes interactions and creates and delivers value to targeted customers to improve customer satisfaction and customer retention at a profit. It is grounded in high quality customer data and enabled by information technology. CRM Systems: Technological part of CRM. Computers and all other information technologies used to help CRM are included. Data Mining (DM): The process of searching and analyzing data in order to find latent but potentially valuable information and to identify patterns and establish relationships from a huge database. Electronic Commerce (E-Commerce): Any business done electronically. The electronic business where information technology is applied to all aspects of company’s operations.

Venkatraman, N. (1989). The concept of fit in strategy research. Academy of Management Review, 14(3), 423444.

Knowledge Discovery in Databases (KDD): The process of data selection, sampling, pre-processing, cleaning, transformation, dimension reduction, analysis, visualization, and evaluation for the purpose of finding hidden knowledge from massive databases.

Wehmeyer, K. (2005). Aligning IT and marketing—The impact of database marketing and CRM. Journal of Database Marketing & Customer Strategy Management, 12(3), 243256.

Online Analytical Processing (OLAP): Various types of online query-driven analyses for examining stored data. OLAP enables a user to easily and selectively extract and view data from different points-of-view.

907

C