intTypePromotion=1 Tuyển sinh 2023 dành cho Gen-Z

Application of Knowledge Management Technology in Customer Relationship Management

Chia sẻ: Monkey68 Monkey68 | Ngày: | Loại File: PDF | Số trang:15

lượt xem
  Download Vui lòng tải xuống để xem tài liệu đầy đủ

Given the important role being played by knowledge management (KM) systems in the current customer-centric business environment, there is a lack of a simple and overall framework to integrate the traditional customer relationship management (CRM) functionalities with the management and application of the customer-related knowledge, particularly in the context of marketing decisions. While KM systems manage an organization’s knowledge through the process of creating, structuring, disseminating and applying knowledge to enhance organizational performance and create value, traditional CRM have focused on the transactional exchanges to manage customer interactions. True CRM is possible only by integrating them with KM systems to create knowledge-enabled CRM processes that allow companies to......

Chủ đề:

Nội dung Text: Application of Knowledge Management Technology in Customer Relationship Management

  1. Knowledge and Process Management Volume 10 Number 1 pp 3–17 (2003) Published online in Wiley InterScience ( DOI: 10.1002/kpm.163 & Research Article Application of Knowledge Management Technology in Customer Relationship Management Ranjit Bose1* and Vijayan Sugumaran2 1 Anderson School of Management, University of New Mexico, USA 2 School of Business Administration, Oakland University, USA Given the important role being played by knowledge management (KM) systems in the current customer-centric business environment, there is a lack of a simple and overall framework to integrate the traditional customer relationship management (CRM) functionalities with the management and application of the customer-related knowledge, particularly in the context of marketing decisions. While KM systems manage an organization’s knowledge through the process of creating, structuring, disseminating and applying knowledge to enhance orga- nizational performance and create value, traditional CRM have focused on the transactional exchanges to manage customer interactions. True CRM is possible only by integrating them with KM systems to create knowledge-enabled CRM processes that allow companies to eval- uate key business measures such as customer satisfaction, customer profitability, or customer loyalty to support their business decisions. Such systems will help marketers address customer needs based on what the marketers know about their customers, rather than on a mass general- ization of the characteristics of customers. We address this issue in this paper by proposing an integrated framework for CRM through the application of knowledge management technology. The framework can be the basis for enhancing CRM development. Copyright # 2003 John Wiley & Sons, Ltd. INTRODUCTION One CRM trailblazer was the gaming company Harrah’s Entertainment, which has successfully CRM is one of the hottest tools in business combined software and human marketing exper- today. But like total quality management and re- tise to get gamblers into its 25 casinos. Harrah’s engineering before it, CRM has not always lived do a thorough, sophisticated analysis of 24 million up to its hype (Brown, 2000; Swift, 2001). Still, com- customers in their database. Harrah’s know—how panies ignore it at the risk of being left behind. frequently customers come, what they play, and Simply, CRM is a high-tech way of gathering mou- they then provide follow-up with continuous com- ntains of information about customers, then using munication over the phone, direct mail and e-mail it to make customers happy—or at least a source and on their Web site. It allows Harrah’s to be par- of more business. It is therefore, concerned with ticipatory rather than being simply reactive. Their understanding and influencing customer behavior technologists refer to it as CRM but their managers (Kotler, 2000). refer it as their loyalty program. Although CRM is the fastest-growing business tool satisfaction with its use currently ranks quite *Correspondence to: Dr Ranjit Bose, Anderson School of Man- agement, University of New Mexico, Albuquerque, NM 87131, low (Winer, 2001). Many companies have started USA. Email: to realize that they need both the mountains of Copyright # 2003 John Wiley & Sons, Ltd.
  2. RESEARCH ARTICLE Knowledge and Process Management information on millions of customers as well as an larly in the context of marketing decisions (Helmke appropriate technical infrastructure coupled with et al., 2001; Massey et al., 2001; Parasuram and Gre- marketing expertise to use CRM satisfactorily wal, 2000). While KM systems manage an organiza- (Zeithaml, 2001). CRM is not necessarily about tion’s knowledge through the process of creating, automating or speeding up existing operational structuring, disseminating and applying knowledge processes; rather, it is about developing and opti- to enhance organizational performance and create mizing methodologies to intelligently manage cus- value (Alavi and Leidner, 2001; Davenport and tomer relationships. Thus, it is about effectively Prusak, 1998; Liebowitz, 1999; Offsey, 1997), tradi- managing and leveraging customer related infor- tional CRM have focused on the transactional mation or knowledge, to better understand and exchanges to manage customer interactions. True serve customers. CRM is possible only by integrating them with A true CRM solution design requires a complex KM systems to create knowledge-enabled CRM pro- combination of many best-of-breed components, cesses that allow companies to evaluate key busi- including analytical tools, campaign management, ness measures such as customer satisfaction, and event triggers, combined with the many new customer profitability, or customer loyalty to sup- components such as collateral management, rule- port their business decisions (Fahey, 2001; Reich- based workflow management, and integrated chan- held and Schefter, 2000; Winer, 2001). Such nel management needed to achieve a one-to-one systems will help marketers address customer marketing capability. This capability dictates the needs based on what the marketers know about need for a single, unified, and comprehensive their customers, rather than on a mass generaliza- view of customers’ needs and preferences across tion of the characteristics of customers. all business functions, points of interactions, and We address this issue in this paper by presenting audiences (Shoemaker, 2001; Tiwana, 2001). Addi- an integrated framework for CRM through the tionally, it requires the existence of interfaces application of knowledge management technology. between non-customer contact systems, such as The framework is designed to deliver consistent ser- enterprise resource planning systems (ERP), and vice across all touch points and channels by provid- operational and customer contact systems. ing: (a) a single view of each customer across the As organizations move towards a comprehensive entire enterprise and throughout the customer’s life- e-business environment, the business processes sup- cycle; and (b) an architecture that supports and pro- porting the environment become increasingly, motes knowledge-based, analysis-driven interaction highly knowledge-intensive and therefore, an orga- with each customer. To test the operational feasibil- nization’s long-term success and growth become ity of this framework, a proof-of-concept prototype dependent on the successful expansion, use, and has been developed and tested that uses current management of its corporate knowledge across its technologies such as extensible markup language business processes (Davenport and Grover, 2001; (XML) and intelligent software agents for perform- Liebowitz, 2000). CRM is no exception to this trend, ing the proposed KM and CRM activities. it is moving away from being a transaction-oriented, Our paper is further organized as follows. First, operational system of the past to a more knowledge- we present a background literature review on oriented, analytical system of the future that pro- CRM, KM and discuss the uniqueness of our vides the means by which a company can maintain work. We then provide the KM capabilities needed a progressive relationship with a customer across for CRM and the architecture for KM-based CRM. that customer’s lifetime relationship with the com- The proof of concept prototype implementation pany (Gordon, 1998; Kalakota and Robinson, and a demonstration session is then presented. Dis- 2001). This means having the ability to track and cussion on the implications as well as limitations of analyze a range of customer actions and events our research and the future research needs are fol- over time, using the information and knowledge lowed by the concluding remarks. from operational CRM systems as well as from other enterprise systems such as KM systems (Wiig, 1999). Given the important role being played by KM BACKGROUND systems in the current customer-centric environ- Customer Relationship Management ment, there is a need for a simple and integrated framework for the management of customer know- CRM is about managing customer knowledge to ledge (Winer, 2001). Surprisingly, there is a lack of a better understand and serve them. It is an umbrella simple and comprehensive framework to integrate concept that places the customer at the center of an the traditional CRM functionalities with the man- organization. Customer service is an important agement and application of the knowledge, particu- component of CRM, however CRM is also 4 R. Bose and V. Sugumaran
  3. Knowledge and Process Management RESEARCH ARTICLE concerned with coordinating customer relations prise, companies can make use of the abilities of across all business functions, points of interaction, analytical CRM systems, and with them, make and audiences (Brown, 2000; Day, 2000). truly customer-centric business decisions. For Delivering consistent service across all touch example, companies can proactively offer products points gives companies a strong market advantage. and services that fit a given customer’s needs based When information or knowledge is fragmented on what the customer has already purchased, or within a company, customer feedback is hard to increase purchase rates by dynamically personaliz- obtain. As a result, customer service suffers and ing content based on Web visitor’s profile, or pro- organizations fall back on the mass marketing prin- vide customers in the highest value tier with ciple that ‘one-size-fits-all’. One-to-one marketing personal representatives who understand their his- requires a comprehensive view of customers’ needs tory or preferences. and preferences (Kotler, 2000). There is an increased sense of urgency in the Information technology-driven relationship institutionalization of comprehensive knowledge management by a firm focuses on obtaining management programs due to the fact that the Inter- detailed knowledge about a customer’s behavior, net and the World Wide Web are revolutionizing preferences, needs, and buying patterns and on the way enterprises do business (Alavi and Leidner, using that knowledge to set prices, negotiate terms, 1999; Leebaert, 1998; Liebowitz, 2000; O’Leary, tailor promotions, add product features, and other- 1998). A well-designed KM infrastructure makes it wise customize its entire relationship with each easier for people to share knowledge during pro- customer (Kohli, 2001; Shoemaker, 2001). Offering blem solving resulting in reduced operating cost, customers convenience, personalization and excel- improved staff productivity, cost avoidance, and lent service plays a key role in the success and dif- soft benefits such as increasing the knowledge ferentiation of many online businesses (Kalakota base, and sharing expertise (Applehans et al., 1999). and Robinson, 2001). CRM focuses on providing The KM framework we present (shown in and maintaining quality service for customers by Figure 1) consists of the following four major pro- effectively communicating and delivering pro- cesses: (a) knowledge identification & generation, ducts, services, information and solutions to (b) knowledge codification & storage, (c) knowledge address customer problems, wants and needs. distribution, and d) knowledge utilization & feed- back. The knowledge identification & generation process includes recognition and creation of new knowledge. Knowledge management It focuses on determining the relevant customer, pro- KM is management of a company’s corporate cess and domain knowledge needed to successfully knowledge and information assets to provide this carry out CRM activities and acquiring or generating knowledge to as many company staff members as this knowledge by monitoring the activities of custo- possible as well as its business processes to encou- mers and other players in the industry. rage better and more consistent decision-making The knowledge codification & storage process invol- (Probst et al., 2000). By integrating operational ves converting knowledge into machine-readable CRM data with knowledge from around the enter- form and storing it for future use. In particular, it Figure 1. Knowledge management framework KM Technology in Customer Relationship Management 5
  4. RESEARCH ARTICLE Knowledge and Process Management deals with archiving the new knowledge by adding retrieved information, filtering away irrelevant it to a persistent knowledge repository that can be and unwanted information, and integrating infor- used by all the stakeholders. This process consists mation from heterogeneous information sources. of mapping the knowledge to appropriate formal- In order to execute tasks on behalf of a business isms, converting it to the internal representation process, computer application, or an individual, and storing it in the knowledge repository. Current agents are designed to be goal driven, i.e. they technologies such as XML and the Universal are capable of creating an agenda of goals to be Description, Discovery and Integration (UDDI) satisfied. Agents can be thought of as intelligent formalism can be used for internal representation computerized assistants. and storage. These approaches facilitate easy search and retrieval of relevant knowledge from the repo- XML and KM sitories, and enables the stakeholders to apply this Extensible Markup Language or XML is emerging knowledge in decision-making (David, 1999). as a fundamental enabling technology for content The knowledge distribution process relates to dis- management and application integration (Balasu- seminating knowledge throughout the organiza- bramanian and Bashian, 1998; Goldfarb and tion and handling requests for specific knowledge Prescod, 1998). XML is a set of rules for defining elements that would be useful in working through data structures and thus making it possible for key a specific problem scenario. Knowledge dissemina- elements in a document to be characterized accord- tion can employ either ‘push’ or ‘pull’ technologies ing to meaning. XML has several valuable character- depending upon the organization’s culture and istics. First, it is a descriptive markup language infrastructure. rather than a procedural markup language. Hence, The knowledge utilization & feedback process com- it is possible to represent the semantics of an XML prises knowledge deployment and providing feed- document in a straightforward way. Second, it is back. This process enables the stakeholders to vendor independent and therefore highly transpor- identify and retrieve relevant knowledge needed table between different platforms and systems while for solving a particular problem. Utilization of this maintaining data integrity. Third, it is human legi- knowledge in the context of a specific problem ble. It is also worth noting that XML has its roots may result in additional knowledge, which can be in SGML (Standard Generalized Markup Language) abstracted out and stored in the knowledge reposi- and adheres to many of its principles. tory for future use. Stakeholders can provide feed- XML enables us to build a structure around the back regarding the quality of knowledge stored in document’s attributes, and RDF (Resource Descrip- the repository as well as how easy or difficult it is tion Framework) allows us to improve search to search for relevant knowledge. They can also iden- mechanisms using the semantics of annotations tify new types of knowledge that need to be gathered (Decker et al., 2000; Rabarijaona et al., 2000). XML based on strategic objectives and the changes that are makes it possible to deliver information to agents taking place within the environment. in a form that allows for automatic processing after This research attempts to integrate relevant receipt and therefore distribute the processing load enabling technologies (Devedic, 1999; Fowler, 2000; over a federation of agents that work cooperatively Sycara et al., 1996; Wu, 2001) into an environment in problem solving. The set of elements, attributes, that would support organizational knowledge crea- and entities that are defined within an XML docu- tion, use, and management. Two such enabling tech- ment can be formally defined in a document type nologies that we discuss are intelligent agents and definition (DTD). XML, which are briefly discussed below. We contend that by combining intelligent agent and XML technologies, one could envision a Intelligent agents and KM knowledge management environment that sup- Intelligent agents are useful in automating repeti- ports all phases of the knowledge life cycle, tive tasks, finding and filtering information, and namely, creation, organization, formalization, dis- intelligently summarizing complex data (Murch tribution, application, and evolution. and Johnson, 1999). Just like their human counter- parts, intelligent agents can have the capability to Our contribution learn and even make recommendations regarding a particular course of action (Hess et al., 2000; We present an integrated framework, that aims for Maes et al., 1999). Intelligent agents can act on knowledge-enabled CRM processes, and which sup- behalf of human users to perform laborious and ports and promotes consistent, knowledge-based, routine tasks such as locating and accessing neces- analysis-driven interaction with each customer. Maj- sary information, resolving inconsistencies in the ority of today’s CRM systems are focused primarily 6 R. Bose and V. Sugumaran
  5. Knowledge and Process Management RESEARCH ARTICLE on call centers’ operations (Brown, 2000; Massey several categories of knowledge management cap- et al., 2001; Orzec, 1998). Several software vendors abilities through the deployment and integration of are active in this field and are offering initial ver- currently available technologies (Gold et al., 2001). sions of their products. Examples include Macrome- The capabilities prescribed in this research are pri- dia (ARIA and LikeMinds product lines), Vignette, marily intranet and extranet based. Engage, IBM (i.e. net commerce), Mathlogic, Micro- The capabilities framework, presented in soft (i.e. Site Server Commerce), NetGenesis, and E. Figure 2, is designed around enterprise knowledge piphany. The analytical CRM system that we pro- portals. Using a portal architecture allows for a pose is just emerging (Swift, 2001). It is designed common interface to knowledge from different to provide business intelligence by encompassing knowledge sources such as documents, applica- knowledge management practices and by lever- tions, and data warehouses (Applehans et al., aging the knowledge gathered from cross-functional 1999; Caldwell et al., 2000). The capabilities frame- customer touch points such as call center, Web work is designed to accelerate the penetration of access, e-mail, and direct sales. knowledge management within organizations The ability to leverage the knowledge from because the users, who most likely are familiar customer-facing systems for back-office analysis with the portal concept through the use of Internet has recently been proven to be directly propor- portals such as Yahoo, will expect that the interface tional to a company’s success in enhancing custo- component of the architecture to offer similar cap- mer loyalty (Reichheld and Schefter, 2000). abilities for knowledge management, such as Without this ability, the environment remains dis- search engines and automatic document summari- connected, and many important business questions zation, across an enterprise-wide collection of cannot be easily answered. For example, a custo- documents. mer service representative scheduling a follow-up At a high level the framework can be explained communication with a customer may not be able as comprised of two parts. First, it is designed to to discern that customer’s value score to determine leverage existing knowledge and to enable creation the level of service that should be provided, or an of new knowledge through a continuous learning account representative may have no idea whether a process denoted by the knowledge learning loops. key business customer has responded to certain And second, the rectangular labeled boxes denote key promotions, or a customer support analyst the KM capabilities and a few currently available may try in vain to measure complaint history techniques or technologies that can provide them. against sales revenue for a given product. A brief description of each of the capabilities is pro- Analytical CRM systems can incorporate several vided below. different types of analytical tools for support per- Presentation involves personalizing both the sonnel. For example, tools for predictive modeling access to and displaying of the results of user inter- (e.g. behavior prediction uses historical customer actions with the system. It is designed to let every behaviors to foresee future behaviors, using sophis- organizational user know where to go to find the ticated modeling and data mining techniques) to organization’s knowledge through a single provide lists of customers most likely to respond browser-based point of entry to all information to a given marketing campaign, or purchase-pattern that the user may need. Personalization provides recognition, or enabling marketing and sales staff to the ability to customize what types of information compare customers with like behaviors so they can are relevant to a user and how that information is position new products to an optimal audience presented. (Berry and Linhoff, 1997; Bose and Sugumaran, The personalization function helps personalize 1999; Fraternali, 1999). The keys to different types content and services to deliver tailored content or of analyses, and especially to the actions that result, information to users based on several user criteria are (a) knowing a firm’s best customers and its or preferences. The primary capabilities of this unprofitable customers, so it can lure the right function include the creation of personalization ones back, and (b) understanding that CRM has to profiles of individual users or groups or depart- work for customers, not just the company. ments or divisions, providing personalized naviga- tion, providing personalized notification, and the ability to personalize the content categorization. KNOWLEDGE MANAGEMENT Personalization is often accomplished by using CAPABILITIES NEEDED FOR CRM software agents, commonly called spiders, to get the information and handle user profiling. In order to implement knowledge-enabled CRM The collaboration function is designed to connect processes, companies need to provide and support people with people through communities of KM Technology in Customer Relationship Management 7
  6. RESEARCH ARTICLE Knowledge and Process Management Figure 2. Knowledge management capabilities for CRM practices; to preserve discussions; and to stimulate are designed in such a way that users can set up collaboration by integrating the knowledge reposi- and control them. The users can specify in them tories and collaboration applications such as work- the type of knowledge he or she wants to publish, flow. distribute, and receive. The frequency (by time The process function allows users to participate and/or quantity) and method (by e-mail or Web in relevant business processes in the context of page) are important parameters that should be set their own roles. Through this function, users have up by the users. access to knowledge management applications The integrated search function is designed to such as knowledge or evidence based decision sup- reduce the information overload and usefulness port system applications that enable increased of search results to the users. Integrated searches responsiveness to customers and partners. across all repositories are performed by default The publishing and distribution function provides but users can also identify the repositories they the means and a platform for users to easily cap- want to search such as Web pages, e-mails, and dis- ture and distribute the particular kinds of knowl- cussions. This function should also provide the edge assets they need to monitor without ability to automate indexing and to crawl fre- requiring them to learn complex programming quently to keep the index current. syntax. Software agents are used extensively for The categorization function allows users to this function (Aguirre et al., 2001). These agents browse, create, and manage knowledge categories. 8 R. Bose and V. Sugumaran
  7. Knowledge and Process Management RESEARCH ARTICLE It establishes a process and guidelines for author- elements from various sources, codifying, storing ing and publishing knowledge categories by the and disseminating knowledge, and utilizing this users. Business groups or departments or divisions knowledge in problem solving (Nissen et al., are made responsible for creating and managing 2000). Hence, we contend that a KM-based CRM their own subject area taxonomies. system would provide precisely the kinds of cap- The integration function ensures seamless and abilities needed for a CRM system to be effective consistent navigation among and between the in managing lasting partnerships with valuable above functions and knowledge sources such that customers. We envision a KM-based CRM system all individuals can use the organization’s combined with components that facilitate the easy gathering knowledge and experience in the context of their and assimilation of customer related information own roles. as well as organizational processes and industry practices. We propose an architecture for a custo- mer centric CRM system, shown in Figure 3, that ARCHITECTURE FOR KM-BASED combines the traditional knowledge management CRM SYSTEMS capabilities as well as the CRM activities needed for successful CRM initiatives. The proposed archi- CRM projects usually fail because they force a lot of tecture consists of four major components: (a) inter- changes quickly on business units and the resulting nal and external data sources, (b) knowledge applications often don’t serve customers any better. acquisition, (c) knowledge repositories, and (d) They also fail to integrate the disparate data knowledge utilization. These components are sources or provide the right kind of information briefly described in the following paragraphs. to the right people at the right time (Parasuram and Grewal, 2000). Hence, CRM applications (a) Data sources: Effective customer relationship should have the capability to not only gather and management requires different types of infor- make available relevant information in a timely mation from a variety of sources. For example, fashion, but also provide tools for analyzing and transaction information may be contained in sharing the information in a meaningful way and operational databases, whereas standard oper- allow managers to act quickly. Knowledge man- ating procedures may be stored in official docu- agement systems deal with these kinds of issues, ments. Data sources may be both internal and particularly, identifying and creating knowledge external to the organization and the CRM Figure 3. KM-based CRM analytics system architecture KM Technology in Customer Relationship Management 9
  8. RESEARCH ARTICLE Knowledge and Process Management system should have mechanisms to access and organizational processes, policies and proce- retrieve relevant data. For example, the CRM dures that have been established and their system should be capable of gaining access to applicability to different situations. Mostly, not only transaction and customer related infor- standard operating procedures are described mation, but also organizational processes and in documents, which are not readily accessible industrywide domain information that would to users. This agent creates a repository of these be useful in problem solving and strategic deci- processes and policies for everyone to access. sion making activities. Thus, the CRM system The Industry Info Agent is structured to access should have an open architecture that is cap- data sources outside the organization to gain an able of interacting with a wide variety of data understanding of the latest developments that and knowledge sources. are taking place in the industry and making Data needed for CRM analytics is very this knowledge available to decision makers. diverse and may be unstructured and difficult (c) Knowledge repositories: This component con- to manage (e.g. emails, call reports on PDAs sists of repositories that contain knowledge ele- etc.). Emerging information technologies can ments generated by humans as well as the bridge the gap by: (a) defining standard data agents that are part of the knowledge acquisi- formats, such as XML for data presentation or tion component. These repositories are continu- Open Database Connectivity for database-to- ally updated as new information becomes database exchanges, (b) ensuring data integrity available. There are four major repositories through proven and published processes, (c) that are maintained, namely, (a) Customer establishing data migration processes, such as Transactions, (b) Customer Profiles, (c) Policies storing procedures for graphical data, and (d) and Procedures, and (d) Domain Knowledge. choosing CRM analytics tools that support The Customer Transaction repository contains Web browser access. particulars about all the transactions related to (b) Knowledge acquisition component: This compo- customers. For each purchasing transaction, nent is responsible for the early phases of information about the products and services knowledge management life cycle, which that the customer bought, discounts that were involves identifying, acquiring and storing provided, date of purchase, etc. are maintained relevant knowledge that would be useful in so that the customer representative can search managing customers and products and making and retrieve one or more transaction records meaningful decisions regarding customer ser- for a particular customer. The Customer Pro- vice and product service offerings. For exam- files repository contains the complete back- ple, keeping track of customer histories and ground of each customer including customer characteristics would be essential in determin- history and preferences. It also contains custo- ing who, and how best to serve the cliental mer ratings and as a result a service representa- given various options. The knowledge acquisi- tive can quickly assess the value of a particular tion component consists of different agents that customer while interacting with that customer, are geared towards acquiring and synthesizing and make appropriate decisions based on the information related to various aspects of custo- importance of the customer. The Policies and mer relationship management. These agents Procedures repository contains information are: (1) Transaction Info Agent, (2) Customer regarding standard procedures and policies Info Agent, (3) Process Info Agent, and (4) that have to be followed in handling a particu- Industry Info Agent. The Transaction Info lar situation. It also contains taxonomies of pro- Agent is responsible for gathering and assimi- duct codes and associated services. The Domain lating information regarding what products a Knowledge repository contains information particular customer has bought over a period about the industry in general, and the latest of time. This information is obtained by inter- developments and trends within that industry acting with the transaction databases that exist that decision makers have to be aware of, within the organization. The Customer Info such as changes in governmental regulations, Agent gathers information related to customer new standards and benchmarks, etc. preferences and characteristics and keeps track (d) Knowledge utilization component: The knowl- of customer profiles. It is primarily responsible edge utilization component is responsible for for generating a comprehensive picture of supporting the later phases of the KM life cycle, every customer and determining the value of in particular, activities related to searching and each customer. The Process Info Agent deals retrieving relevant knowledge, as well as shar- with collecting information related to various ing this knowledge with other stakeholders to 10 R. Bose and V. Sugumaran
  9. Knowledge and Process Management RESEARCH ARTICLE be utilized in different scenarios. It acts as the vide better profit for the space? Without this interface to knowledge repositories. It enables capability, store managers may have no way stakeholders to search the knowledge reposi- of identifying the most profitable products tories for specific information related to the and allocating more time to these profitable problem they are solving. This component is lines. also responsible for content delivery (knowl- (iii) Predictive Modeling Agent: On the CRM analy- edge that may be of interest to certain groups) tics side, the biggest disappointment has been on a periodic basis. The knowledge utilization the failure to integrate business logic into the component consists of the following agents: (i) tools. The Predictive Modeling Agent enable Repository Management Agent, (ii) Situation managers to conduct meta-analysis and identi- Analysis Agent, (iii) Predictive Modeling Agent, fy areas of strengths and weaknesses. For and (iv) Marketing Automation Agent. example, they can watch transactions in real time to spot patterns, such as decreasing trans- (i) Repository Management Agent: This agent pro- action rates or balances for a high value custo- vides a number of functions for repository mer that indicate that a customer might soon management such as organizing, maintaining leave. It enables managers to get a grasp on and evolving the knowledge repositories. It customers’ buying patterns, anticipate trends also provides mechanisms for browsing these and more carefully align inventory to maxi- repositories as well as searching for specific mize profits in a chain of stores. Most timely knowledge elements relevant to a particular information is of little use unless the corporate problem at hand. This agent is also responsible strategy aligns with what the customer data is for knowledge dissemination, which includes revealing. various aspects such as presentation, persona- (iv) Marketing Automation Agent: One of the big- lization, collaboration, and publishing. This gest pitfalls of customer databases is that the agent provides easy access to important and best customers are bothered endlessly—sur- relevant data, in particular, makes more custo- veys, new offers, cross-selling etc. Lack of an mer data available to call center operators so integrated CRM system results in alienating they can solve customer problems on the first customers by making inappropriate pitches call. This agent disseminates the information and ignoring customers with low current mined by analysts to the marketing, sales, returns but high potential. Another mistake and front-line customer service people who that is often made is segmenting customers could actually use it. It also permits caller on the basis of demographics such as age, identification linked with customer histories income, sex or education because this informa- and characteristics in order to identify most tion is relatively easy to get. But the best CRM valuable customers and provide appropriate systems will segment customers based on fun- services. damental values. The proposed KM-based (ii) Situation Analysis Agent: This agent provides CRM system will be able to match actual buy- mechanisms for the user to undertake problem ing information to customer profiles and pre- solving and decision-making activities. For ferences, which can permit the marketing example, a customer service representative people to really see trends from individual may be faced with an angry customer with a customers and develop better marketing cam- complaint. The representative can analyze the paigns. situation and reach a resolution quickly based on the customer profile and transaction his- tory. Similarly, a manager has the ability to see which specific products in the store are selling well, badly or according to expected PROTOTYPE IMPLEMENTATION trends, and to take appropriate actions. The manager would have the capability to ask sev- A proof-of-concept prototype is currently under eral key questions such as: is the product per- development. This prototype uses the traditional forming badly because of poor display client-server architecture, where the client is a sim- standards, poor stock availability or incorrect ple web browser, using which the user can interact location? Is the product right for the store, with the knowledge repositories. The user can also does it provide enough profit for the space perform one or more CRM activities supported by allocated, could another product’s space be the Knowledge Utilization component. The agents enlarged or a new product brought in to pro- that are part of the Knowledge Acquisition KM Technology in Customer Relationship Management 11
  10. RESEARCH ARTICLE Knowledge and Process Management Component as well as the Knowledge Utilization data directly in native XML format and provide Component have been implemented using JADE facilities for fast storage, exchange and retrieval of (Java Agent DEvelopment Framework) from XML documents. CSELT, Turin, Italy (Bellifemine et al., 1999). Sample session with prototype: The following JADE is a middle-ware product that is used to paragraphs describe a brief sample session that develop agent-based applications, which are in provides a glimpse of some of the functionalities compliance with the FIPA specifications for intero- of the KM-based CRM System prototype. When perable intelligent multi-agent systems. JADE is the user accesses the CRM system, a login screen, java-based and provides the infrastructure for shown in Figure 4, is presented where the user agent communication in distributed environments, can type in the userid, password and the user based on FIPA standards. The reasoning capability type. Users are provided different levels of access of the agents has been implemented through JESS, to control the evolution of the knowledge reposi- which is an expert system shell written in Java tories. For example, not all users can create new (Friedman-Hill, 2002). The transaction information, knowledge elements and store them in the reposi- customer profiles and preferences, organizational tory or have access to sensitive information. Some processes and procedure information, as well as of the typical users of the system are customer ser- the application domain knowledge are captured vice representatives, department heads, division and represented in XML documents with appropri- managers and senior executives. Once the user is ate DTDs. These XML documents are stored in the authenticated, depending upon the type of the corresponding knowledge repositories, which have user, appropriate menus are presented. Users can been implemented as XML databases using the also customize the interface to suite their tastes Tamino software (from Software AG— and preferences. Among other things, Tami- When the user logs into the system, he or she can no provides X-Studio, which is a complete suite of perform various knowledge management and application development tools for creating XML- CRM activities. For example, if the user type is based applications. Tamino XML databases store ‘customer service representative,’ he/she can, Figure 4. Initial screens from the KM-based CRM system 12 R. Bose and V. Sugumaran
  11. Knowledge and Process Management RESEARCH ARTICLE among other things, view customer profiles and of a particular customer and get a sense of the histories as well as perform situation analysis. Fig- value and loyalty of that customer. The customer ure 4 shows the initial menu that lists the options transaction knowledge repository has been imple- for carrying out various functions. The user can mented as an XML database and can be searched select any of the options and click on the Submit using customer id or customer name. Figure 5 button to perform that particular operation. The shows the interface for searching and viewing ‘Knowledge Acquisition and Repository Manage- transaction histories. By default, the system dis- ment’ option enables the user to invoke the knowl- plays the transaction history of a customer as an edge elicitation process from various sources or XML document (bottom portion of Figure 5), and perform maintenance operation on one or more of by clicking on the Display button, the user can the knowledge repositories. The user can explicitly see the HTML rendering of the document using specify tasks for the agents that are part of the cascading style sheets. knowledge acquisition component, or ask them to The user can also browse the organizational pro- gather information on a continual basis. These cesses and procedures knowledge repository or agents can create new knowledge elements and search for specific policies related to a particular add them to the appropriate knowledge repository situation by selecting the ‘Policies and Procedures’ using predefined ‘repository management’ proce- option shown in Figure 4. The system provides dures. The ‘Customer History and Profile’ option another panel where the user can specify a few key- facilitates the user to probe available customer words, using which the repository is searched and information and generate an up-to-date picture of matching policies and procedures are displayed to a particular customer and determine the value of the user. The ‘Situation Analysis’ option enables that customer. For example, a customer service the user to analyze a particular event or circum- representative can pull up the transaction history stance based on relevant information, and perform Figure 5. Viewing transaction information for a specific customer KM Technology in Customer Relationship Management 13
  12. RESEARCH ARTICLE Knowledge and Process Management Figure 6. Situation analysis interface what-if analysis before reaching a meaningful con- search the transaction and customer profile reposi- clusion. For example, a customer service represen- tories. The retrieved information is displayed in the tative may be faced with a situation where a window shown in the lower part of Figure 6. Simi- customer is not satisfied with a product or service larly, when the user enters situation descriptors, and is calling up and demanding recourse. In this those keywords are used in searching the policies situation, having quick access to that customer‘s and procedures repository and relevant policy history and rating, as well as the policies and pro- and procedure information is displayed in the low- cedures that dictate how such a case should be er window. handled, can help that representative quickly The ‘Predictive Modeling’ and ‘Marketing Auto- resolve this situation to the satisfaction of the cus- mation’ options (shown in Figure 4) are utilized by tomer and still stay within the parameters that the managers interested in analyzing the performance representative has to operate under. A simple of specific products or services and also under- situation analysis interface is shown in Figure 6. standing the customer base for developing specific The representative can enter customer information marketing campaigns and promotions. The ‘Search in the ‘Customer Info’ box and some keywords Knowledge Repositories’ option (shown in describing the situation in the ‘Situation Info’ box Figure 4) provides ad hoc querying capabilities and get relevant customer information as well as using which the user can search one or more applicable policies and procedures displayed by knowledge repositories for related information. clicking on the appropriate buttons. When the user clicks on the ‘Recommendation’ button, the system provides some recommendations based on DISCUSSION pre-established rules. When the user clicks on the customer profile button after entering the customer This section highlights some of the major implica- identifier, an appropriate query is generated to tions of this research. Our approach to integrating 14 R. Bose and V. Sugumaran
  13. Knowledge and Process Management RESEARCH ARTICLE knowledge management techniques into customer proposed integration framework, it is by no means relationship management activities provides a full-blown system. Therefore, the current version several advantages. Individuals, various business of the prototype has the following limitations. First, units, and the organization as a whole can all ben- while we have developed the DTDs for some of the efit from the proposed integrated KM-based CRM common knowledge elements, much work remains environment. At the individual level, customer ser- to be done in order to capture all types of knowl- vice representatives can browse the knowledge edge that would be useful in carrying out repositories, perform plain-text searches for speci- comprehensive CRM activities. Second, knowledge fic customer information, customer profile and his- repositories as well as operating procedures evolve tory, and rating. This real time access to relevant over time in an organizational setting. Hence, information enables the representatives to better the prototype needs additional capability to ensure serve customers. Different business units can bene- that the knowledge repositories are consistent with fit from such a system by being able to gain access business processes on a continual basis. Third, the to customer and sales information that are gathered prototype currently does not provide application through various touch points, as well as the stan- interface to several potential third party software dard policies and procedures that are otherwise that could be easily utilized in predictive modeling not easily accessible. At the organizational level, and automating many of the marketing related our system could be utilized in providing a com- activities. mon infrastructure for carrying out customer Further work is required to bring the prototype relationship management activities and institutio- to a full-blown system as well as to address some of nalizing a comprehensive set of CRM policies. the issues that arise in integrating knowledge man- The proof-of-concept prototype we implemen- agement techniques into customer relationship ted, demonstrates the operational feasibility of the management. Our future work on the prototype proposed KM-CRM integration framework. Cur- includes incorporating additional components for rent technologies such as intelligent agents and knowledge acquisition and utilization, and provid- XML technologies were selected and used for ing APIs for various decision analytic tools for implementation because (1) to reduce the cognitive facilitating the creation of an integrated KM–CRM burden on the user in problem solving and decision portal with customizable functionalities. The making activities, and (2) these technologies facili- resulting full-blown system will be able to support tate the easy integration of knowledge manage- better query facilities for searching knowledge ment activities and CRM activities. For example, repositories, particularly, natural language based intelligent agents can be tasked to monitor certain interfaces that provide flexible query mechanisms. types of transactions or search and retrieve specific Subsequently, field testing and empirical validation customer related information in real time. XML of the full-blown system is necessary to evaluate its technology permits easy codification and dissemi- effectiveness from the perspective of target users. nation of knowledge elements to interested parties While the present prototype uses current technolo- through push or pull technologies. In addition, it gies such as intelligent agents and XML, potential improves the interoperability of knowledge ele- use of other enabling technologies like Ontologies, ments between different applications. Tradit- UDDI (Universal Description, Discovery, and Inte- ionally, KM tools use proprietary knowledge gration), and Web Services need to be investigated. structures and internal representations that prohi- Future research should also address additional bit the exchange of knowledge between various issues related to the integration of KM and CRM applications. In contrast, our system uses XML activities such as configuring KM activities to align representation, which alleviates this problem to a with the overall objectives of CRM initiatives, iden- great extent. Storing customer information in an tifying the types of knowledge needed for specific XML database also facilitates various stakeholders CRM activities, and managing the evolution of a to view information at different levels of aggrega- consistent set of knowledge repositories. tion through specific transformations. For example, customer service representatives can query the XML database for individual customer histories CONCLUSION and profiles, whereas, marketing people can view customer information based on certain ‘value pro- Analytical CRM systems achieve a single, unified positions’. view of the customer and facilitate a seamless While the implemented prototype incorporates exchange between customers and corporations. the necessary functionalities and capabilities to However, a single view of customers requires adequately prove the operational feasibility of the tightly integrated applications both within the KM Technology in Customer Relationship Management 15
  14. RESEARCH ARTICLE Knowledge and Process Management realm of CRM applications and back-end technolo- Decker S, Melnik S, Harmelen FV, Fensel D, Klein M, gies, such as knowledge management. Organiza- Broekstra J, Erdmann, M, Horrocks I. 2000. The seman- tic web: the roles of XML and RDF. IEEE Internet Com- tional knowledge creates value in use. A key puting 4(5): 63–74. challenge in the application of knowledge is trans- Devedzic V. 1999. A survey of modern knowledge mod- ferring it from where it was created or captured to eling techniques. Expert Systems With Applications 17(4): where it is needed and should be used. 275–294. We tried to address the issue in this research by Fahey L. 2001 Linking E-business and operating pro- cesses: the role of knowledge management. IBM Sys- developing a simple and overall framework to inte- tems Journal 40(4): 889–907. grate the traditional CRM functionalities with the Fowler A. 2000. The role of AI-based technology in sup- management and application of knowledge in the port of the knowledge management value activity context of marketing decisions. The operational cycle. Journal of Strategic Information Systems 9(2/3): feasibility of the framework was tested through a 107–128. Fraternali P. 1999. Tools and approaches for developing proof-of-concept prototype, which was built using data-intensive web applications. ACM Computing Sur- intelligent agent and XML technologies. We con- veys 31(3): 227–263. tend that the framework can be the basis for enhan- Friedman-Hill E. 2002. Jess, the expert system shell. cing CRM system functionalities and development. Sandia National Laboratories, Albuquerque, NM, URL: Gold A, Malhotra A, Segars A. 2001. Knowledge manage- ment: an organizational capabilities perspective. Jour- REFERENCES nal of Management Information Systems 18(1): 185–214. Goldfarb CF, Prescod P. 1998. The XML Handbook. Aguirre JL, Brena R, Cantu FJ. 2001. Multiagent-based Prentice Hall: Upper Saddle River, NJ. knowledge networks. Expert systems With Applications Gordon I. 1998. Relationship Marketing: New Strategies, 20(1): 65–75. Techniques to Win the Customers You Want and Keep Alavi M, Leidner DE. 1999. Knowledge management sys- Them Forever. John Wiley: New York. tems: issues, challenges, and benefits. Communications Helmke S, Dangelmaier W, Uebel MF. 2001. CRM- of the AIS 1(7). systems as technology enabler for a customer-oriented Alavi M, Leidner DE. 2001. Review: knowledge manage- knowledge management. PICMET ’01—Portland Inter- ment and knowledge management systems: concep- national Conference on Management of Engineering and tual foundations and research issues. MIS Quarterly Technology 1. 25(1): 107–136. Hess TJ, Rees LP, Rakes T R. 2000. Using autonomous Applehans W, Globe A, Laugero G. 1999. Managing software agents to create the next generation of deci- Knowledge: A Practical Web-Based Approach. Addison- sion support systems. Decision Sciences 31(1): 1–31. Wesley: New York. Kalakota R, Robinson M. 2001. E-Business 2.0: Roadmap Balasubramanian V, Bashian A. 1998. Document manage- for Success. Addison-Wesley: Boston, MA. ment and web technologies: Alice marries the Mad Kohli R. 2001. Managing customer relationships through Hatter. Communications of the ACM 41(7): 107–114. E-business decision support applications: a case of hos- Bellifemine F, Poggi A, Rimassa G. 1999. JADE—A FIPA- pital–physician collaboration. Decision Support Systems compliant agent framework. Proceedings of PAAM’99, 32(2). London, 97–108. Kotler P. 2000. Marketing Management. Prentice-Hall. Berry MJA, Linhoff G. 1997. Data Mining Techniques: For Upper Saddle River, NJ. Marketing, Sales, and Customer Support. John Wiley: Leebaert D. 1998. The Future of the Electronic Marketplace. New York. MIT Press: Boston, MA, Bose R, Sugumaran V. 1999. Application of intelligent Liebowitz J. 1999. Information Technology Management—A agent technology for managerial data analysis and Knowledge Repository. CRC Press: Boca Raton, FL. mining. Database for Advances in Information Systems Liebowitz J. 2000. Building Organizational Intelligence—A 30(1): 77–94. Knowledge Management Primer. CRC Press: Boca Raton, Brown SA. 2000. Customer Relationship Management: A FL. Strategic Imperative in the World of E-Business. John Maes P, Guttman RH, Moukas AG. 1999. Agents that buy Wiley: New York. and sell. Communications of ACM 42(3): 81–87. Caldwell N, Clarkson PJ, Rodgers P, Huxor A. 2000. Massey AP, Montoya-Weiss MM, Holcom K. 2001. Re- Web-based knowledge management for distributed engineering the customer relationship: leveraging design. IEEE Intelligent Systems 15(3): 40–47. knowledge assets at IBM. Decision Support Systems 32: Davenport TH, Grover V. 2001. General perspectives 155–170. on knowledge management: fostering a research agen- Murch R, Johnson T. 1999. Intelligent Software Agents. Pre- da. Journal of Management Information Systems 18(1): ntice Hall: Upper Saddle River, NJ. 5–21. Nissen M, Kamel M, Sengupta K. 2000. Integrated analy- Davenport TH, Prusak L. 1998. Working Knowledge: How sis and design of knowledge systems and processes. Organizations Manage What They Know. Harvard Busi- Information Resources Management Journal 13(1): 24–43. ness School Press: Cambridge, MA. Offsey S. 1997. Knowledge management: linking people David M. 1999. SQL-based XML structure data access. to knowledge for bottom line results. Journal of Knowl- Web Techniques June: 67–72. edge Management 1(2): 113–122. Day GS. 2000. Managing marketing relationships. Journal O’Leary DE. 1998. Enterprise knowledge management. of the Academy of Marketing Science 28(1): 24–31. IEEE Computer March. 16 R. Bose and V. Sugumaran
  15. Knowledge and Process Management RESEARCH ARTICLE Orzec D. 1998. Call centers take to the web. Datamation Sycara K, Pannu A, Williamson M, Zeng D, Decker K. June. 1996. Distributed intelligent agents. IEEE Expert 11(6): Parasuram A, Grewal D. 2000. The impact of technology 36–45. on the quality–value–loyalty chain: a research agenda. Tiwana A. 2001. The Essential Guide to Knowledge Manage- Journal of the Academy of Marketing Science 28(1): 168–175. ment: E-Business and CRM Applications. Prentice Hall: Probst G, Raub S, Romhardt K. 2000. Managing Knowledge: Upper Saddle River, NJ. Building Blocks for Success. John Wiley: Chichester. Wiig KM. 1999. What future knowledge management Rabarijaona A, Dieng R, Corby O, Ouaddari R. 2000. users may expect. Journal of Knowledge Management Building and searching an XML-based corporate mem- 3(2): 155–166. ory. IEEE Intelligent Systems 15(3): 56–63. Winer RS. 2001. A framework for customer relationship Reichheld F, Schefter P. 2000. E-loyalty—your secret management. California Management Review 43(4): 89– weapon on the web. Harvard Business Review July– 107. August. Wu DJ. 2001. Software agents for knowledge mana- Shoemaker ME. 2001. A framework for examining IT- gement: coordination in multi-agent supply chains enabled market relationships. The Journal of Personal and auctions. Expert Systems With Applications 20(1): Selling & Sales Management 21(2): 177–185. 51–64. Swift RS. 2001. Accelerating Customer Relationships: Using Zeithaml VA. 2001. The customer pyramid: creating and CRM and Relationship Technologies. Prentice Hall: Upper serving profitable customers. California Management Saddle River, NJ. Review 43(4): 118–145. KM Technology in Customer Relationship Management 17



Đồng bộ tài khoản