An Intelligent-Agent System Approach in Value Creation Henry W. L. Ho School of Marketing & Mgmt Charles Sturt University
[email protected]
Denise Jarratt School of Marketing & Mgmt Charles Sturt University
[email protected]
Mohamed Anver School of Marketing & Mgmt Charles Sturt University
[email protected]
Abstract Since the early 1990s, increasing turbulence through globalisation and rapidly changing markets have highlighted the need for organisational flexibility. To compete within this environment, organisations have to find ways to enhance value creation for potential and existing customers in an efficient, effective and timely manner. This proposed framework examines how superior customer value can be generated within a business-to-business (B2B) virtual network organisation (VNO) through the application of an Intelligent-Agent System (IAS). It is proposed to integrate the agent system within a virtual network model to create customer value through the integration of agents simultaneously representing customer focus, customer orientation and market orientation. Keywords: Customer value, virtual network organisation, Intelligent-agent system
1.0 Introduction Today’s market environment is described as turbulent due to global competition and rapid changes in industry structure (Achrol 1997; Eggert & Ulaga 2002). To compete within this turbulent market environment, organisations have to find new ways to create value for their potential and existing customers. This research proposes a framework for modelling superior customer value within real-time to serve the rapid changes in today’s business-to-business (B2B) environment.
2.0 Customer Value Research Customer value has proven to be a difficult concept to define and measure (Woodruff 1997; Zeithaml 1998). However, most researchers view customer value as the results or benefits (of goods or services) customers receive in relation to total costs (such as price paid plus other costs related to the benefits) (Christopher 1996; Zeithaml 1998; Mc Naughton, Osborne & Imrie 2002; Kotler, Adam, Brown & Armstrong 2003). An extension of this view, and important to this research, is Woodruff’s (Woodruff 1997) value hierarchy which recognises how the concept of value is reconfigured overtime, and, prior to repurchase considered within the context of value generated during consumption or use, compared to overarching goals. The objective of this research is to develop a framework to examine how superior customer value can be created and delivered to the customer within a virtual, market focused, network organisation in the business-to-business sector (business market). It is proposed that customer value within the business market should be considered as the trade-off between the multiple benefits and sacrifices of a supplier’s offering, as perceived by major decision-makers in the customer’s organisation during consumption and repurchase deliberations, taking into consideration the available alternative supplier’s offering in a specific use situation (Eggert & Ulaga 2002).
3.0 Virtual Network Organisation 1624
Various network organisational (VNO) forms have emerged from the formalisation of relationships between, for example, suppliers, competitors and other vertical or horizontal partners from the early 1990s. Researchers such as Achrol (1997), and Cravens and Piercy (1994) explained new organisational forms as “networks” or, where the network is predominantly managed electronically, “virtual networks”. Researchers have proposed numerous characteristics to describe network organisations (Hale & Whitlam 1997; Franks 1998; Pihkala, Varakami & Vesalainen 1999; Wang 2000; Lau, Wong, Ngai & Hui 2003). Different forms of network organisations can be distinguished based on the objectives and operating principles of the organisation. For the purpose of this research, a virtual network is defined as one that operates predominantly through electronic interaction. That is, the IT architecture links actors, and overarches managerial agreement, with little physical interaction existing across network actors and client organisations.
4.0 Value Creation in Virtual Network Organisations Recent research has identified the cultural, structural and process barriers that face organisations as they strive to retain a customer focus while becoming market oriented (Hadcroft & Jarratt 2004). It has been argued that structures, systems and the culture that. A virtual network, as defined above, facilitates the introduction of intelligent agents that can simultaneously adopt multiple orientations for network engagement with the market and its customers. In the real world, an organisation will adopt an orientation most relevant to its life cycle position and competitive strategy, and build systems and processes to support that orientation. In a virtual world, intelligent systems can deliver value generated in response to systems of agents capturing value requirements and advancements that draw on teams of agents simultaneously representing client organisation needs, client organisation’s customers’ needs, collective learning from customer groups, competitive action, technology changes and other resource changes. One of the major tasks of the VNO is to shorten the business process life-cycle with the objective of generating and delivering better value to business clients and their customers in real time. To achieve real-time value creation within the VNO, intelligent-agents can participate and play an important role in identifying components and forms of value appropriate to achieve customer satisfaction for both client organisations and their customers.
5.0 Intelligent-agent (IA) Research on intelligent-agents or agent-based systems started in the late 1980s (Jennings & Wooldridge, 1998). Following this period, in the early 1990s the focus shifted to the learning capability of agents (Kupfer 1994; Maes 1994). Later research explored the intelligence capability of such agents with a view to their mimicking human actions (Jennings & Wooldridge 1998). In brief, an agent is a program that provides an ideal method to investigate cooperating activities (Jennings & Wooldridge 1998). Additionally, an agent is a program that performs a specific task with a minimum or even without direct human supervision. IAS can combine both computers and humans, working cooperatively over space and time to solve a variety of complex problems. 6.0 Proposed conceptual model of a B2B virtual network organisation 1625
Figure 1 illustrates a VNO containing three network partners and three different business customers. The customer-oriented network theoretical model represented in Figure 1 relies on its internal or external network partners’ resources to generate value laden services for client organisational customers, minimising production costs and producing the value in real time to deliver total customer satisfaction. Intelligent-agents are adopted to monitor the processes, execute any specific tasks given, integrate value attributes and thus ensure that the services and products are of a quality that meets or exceeds customer expectation.
Figure 1: Intelligent-agent in a virtual network organisation
The connections among the network partners are the information and knowledge exchange and the services that are transferred from one to another. Intelligent-agents deployed within this VNO provide the semantic support infrastructure for high-level elements. The CFM (Monitoring agent) monitors and adapts value based on the client organisation needs and feedback. In addition to drawing on feedback accumulated from the CFM, the MOM (Mobile agent) moves from its home server to network partner servers to retrieve reliable client organisation’s customer data and transfer that value relevant data back to home server for integration with the COM (Profiling / Best Practice agent) through the NRCM (Collaborative agent). The COM will be capable of analysis, suggesting best practice solutions and solution changes from data retrieved from the network partners and the CFM. Those best practice solutions will then pass over to the SCVM (Recommending agent) via the NRCM. The SCVM will then have access to the necessary knowledge and real time information to deliver a quality solution to resolve any ad hoc requirements specified by the client organisations or a potential client organisation. Each agent is thus responsible for performing duties that are assigned depending on that agent’s characteristics. A virtual network value solution system including the integration of multiple-agents will have a common infrastructure and agent architecture to support system functioning, specifically, sharing of data, processing resources across networks, exchanging information and collaborating on tasks and goals with the agents representing network orientations and client organisations. 1626
7.0 Integration of Fuzzy Logic into the Agent Driven Organisational Model In the previous sections we proposed a model of a VNO in which an IAS plays a dominant role. That is, the effective functioning of this VNO relies on an automated business planning / execution system, the heart of which constitutes a set of software agents simulating human personnel in a conventional organisation. The success of this IAS working in real / quasi-real time relies heavily on the intelligent behavioural characteristics of the participating softwareagents. Hence, these behavioural characteristics need to be properly defined and programmed into the system in such a way the agents achieve the capability to plan and execute their tasks while optimising the performance of the VNO as a whole. However, due to complex nature of business decision processes and inter-agent communication requirements, designing a set of rules based on rigid conditions of activation, does not render an optimum methodology to define the behavioural characteristics of these agents. Instead, the design process requires an intelligent framework which is flexible and can handle ambiguity and vagueness associated with the decision making processes in typical business environments. Towards this end, we employ Fuzzy Logic (FL) which belongs to the paradigm of Artificial Intelligence (AI), as a tool to define the action-rules of these softwareagents. 7.1 A Brief Introduction to Fuzzy Logic Fuzzy Logic, the core idea of which rests on processing data by defining partial set membership rather than crisp set membership, was founded by Professor Lotfi A. Zadeh in 1965 (Zadeh 1965), at the University of California, Berkley. Professor Zadeh reasoned that we, the humans, do not require precise, numerical information input, and yet we are highly capable of processing these inputs to achieve our objectives. In other words, FL provides a simple way to arrive at a definite conclusion based upon vague, ambiguous and imprecise input information. FL's approach to decision making mimics how a human would make decisions, only much faster. This is achieved by using a set of rules of the form: If x is X and y is Y Then z is Z, rather than attempting to model a system mathematically. The FL model is empirically-based, relying on human experience / expertise rather than a technical / precise understanding of a system. As mentioned above rule, X, Y and Z are fuzzy sets defining the universe of discourse of x, y and z, respectively. To illustrate the concept of a fuzzy set, we take an example variable ‘Cutomer_Demand’ associated with a certain product manufactured by a business and show how it is broken down into a collection of overlapping fuzzy sets, see Figure 2.
Figure 2
1627
The degree of membership indicates to what level a given value for ‘Cutomer_Demand’ represents the overlying semantic class. These classifications are dependent on the actual model, and what constitutes the range of ‘Medium’ in one context might not be valid in another. Fuzzy sets, however, allow us to write rules that apply to multiple semantic contexts at the same time given a single piece of information. 7.2 Fuzzy Logic Rules to define Agent Behaviour In this section, we present some examples of fuzzy logic rules we intend to include in the knowledge base of the proposed intelligent agent framework. It is assumed that agents have access to important information needed for strategic decision making such as customer demographic data, sales figures, profit analysis results etc, that are often available in organisational data warehouses. Further, agents also possess knowledge of trading patterns and demand-supply variations as well as history of customers and preferences, which can be derived from organisation’s databases using intelligent data mining techniques. In addition, financial indicators such as interest rates, exchange rates and stock market prices are fed into the agents’ knowledge repositories via online resources. For the purpose of demonstrating the idea of integrating fuzzy knowledge bases, within restricted space limits, we give below some fuzzy logic based rules as examples. These rules depict how fuzzy set theory and fuzzy if-then rules can augment the strategic decision making capabilities of software agents in a human like manner using imprecise and vague input information. Table 1 explains five different types of variables that can be used within the fuzzy rules and represent the action-rules of the IAS within the VNO: Variables:
Explanation:
Customer_Experience
Customer background, history or past experience of existing client organisations.
New_Solution
New ideas, outcomes or extra value for existing client organisations.
Customer_Desirability
Potential or new client’s interest and determination to do business with us.
Standard_Solution
Standard value or solution based on past experience that can be provided to new client organisations.
Repeat_Value
Standard or ordinary solution that can be provided to existing client organisations. Table 1: Explanation of variable
We give below, a set of example fuzzy rules for each of the three(3) customer situations that could exist in relation to a VNO. Existing client 1 (new solution): If Customer_Experience is HIGH Focus_Priority is HIGH
and New_Solution is LOW
Then
If New_Solution is HIGH Focus_Priority is MEDIUM
and Customer_Experience is LOW
Then
If Customer_Experience is HIGH Focus_Priority is HIGH
and New_Solution is HIGH
Then
1628
(1) If Customer_Experience is LOW Focus_Priority is HIGH
and New_Solution is LOW
Then
and Standard_Solution is LOW
Then
(2) If Standard_Solution is HIGH Focus_Priority is HIGH
and Customer_Desirability is LOW
Then
If Customer_Desirability is HIGH Focus_Priority is HIGH
and Standard_Solution is HIGH
Then
(3) If Customer_Desirability is LOW Focus_Priority is HIGH
and Standard_Solution is Low
Then
New client (standard solution): If Customer_Desirability is HIGH Focus_Priority is HIGH
Existing client 2 (repeat purchase standard solution): If Repeat_Value is HIGH and New_Solution is LOW Focus_Priority is HIGH
Then
If New_Solution is HIGH Focus_Priority is MEDIUM
and Repeat_Value is LOW
Then
If Repeat_Value is HIGH Focus_Priority is HIGH
and New_Solution is HIGH
Then
(4) If Repeat_Value is LOW Focus_Priority is HIGH
and New_Solution is LOW
Then
Taking one of the rules as an example and describing it further, the first rule can be explicitly stated as: If the Customer_Experience has been ‘high’ and New_Solution requirement is ‘low’ then, for example, the Superior Customer Value Manager (Recommending Agent), modifies its behaviour so as to set its ‘Focus_Priority’ (the Recommending Agent’s response towards the respective client) parameter to a high value. Many of the rules are self explanatory, however several require further detail (marked as 1, 2 3 and 4 above). In situations 1 and 4, although we deal less often with this existing client, they rarely request a new service and are even less likely to request a repeat purchase from us. But since this VNO is seek to leverage extra value and to create close relationships with their client organisations. Therefore, ‘Focus_Priority’ for both these situations is classified as high. In a similar manner, in situation 2, although the customer is not a current purchaser of the network’s business solutions, however, if we are capable of provided value for this potential client based on our past experience, then the VNO is seeking to leverage the superior value and to create a close relationship with this potential, new client organisation. Therefore, the ‘Focus_Priority’ is classified as high. Finally in situation 3, the customer’s business solution requirement are not closely aligned with the business solution expertise of the network and thus it would be difficult for the network to tailor-make a solution that perfectly matched this potential client’s request based on our past experience alone. However, through the co-ordination of intelligent agents within the VNO and the support of our network partners, an improved solution can be generated and
1629
superior customer value offer will build a relationship with this new client. Therefore, again, the ‘Focus_Priority’ is set to be high. It is to be noted, as opposed to classical crisp logic, that these decisions are made analogous to a human decision using the fuzzy set theory. 8.0 Future Work The intention of this paper is to provide a brief overview of how superior customer value can be generated within the VNO adopting multiple simultaneous orientations. The essential points relating to value creation within the B2B virtual environment have been briefly introduced. Today’s organisation relies on technology to provide information, facilitate communication and support decision making. Implementation of an IAS within the VNO as shown in figure 1 identifies components and forms of value appropriate to achieve superior value design and delivery for both client organisations and their customers. Further research is being conducted to fully understand how IAS can be used to significantly improve the quality and quantity of the communication and at the same time, shorten the lifecycle of business processes with the objective of generating and delivering improved value to business clients and their customers in real time. References Achrol, R.S. “Changes in the Theory of Interorganisational Relations in Marketing: Toward a Network Paradigm”, Journal of Academy of Marketing Science, Vol. 25, 1997, pp.56-67. Christopher, M. “From Brand Value to Customer Value”, Journal of Marketing Practice, Vol. 2, 1996, p.55. Cravens, D.W. and Piercy, N.F. “Relationship Marketing and Collaborative Networks in Services Organisation’, International Journal of Services Industry Management, Vol. 5, No. 5, 1994, pp.39-53. Eggert, A., Ulaga W. “Customer Perceived Value: A Substitute for Satisfaction in Business Market?”, Journal of Business and Industrial Marketing, Vol. 17, 2002, pp.107-118. Franks, J. “The Virtual Organisation”, Work Study, Vol. 47, No. 4, 1998, pp.130-134. Hale, R. and Whitlam, P. “Towards the Virtual Organisation”, McGraw-Hill, 1997. Hadcroft, P. and Jarratt, D.G. “Market Orientation: an iterative process of customer and market engagement” ANZMAC, Wellington, Nov29-Dec 2, 2004 (forthcoming). Jennings, N.R. and Wooldridge, M. “Applications of Intelligent-agents”, Agent Technology Foundations, Applications and Market, Springer-Verlag, 1998. Kotler, P., Adam, S., Brown, L. and Amstrong, G. Principles of Marketing, Prentice Hall, 2003. Kupfer, A. “Software Agents Will Make Life Easy”, Fortune, Vol. 24, January, 1994, pp.72-73. Lau, H.C.W., Wong, C.W.Y., Pun, K.F. and Chin, K.S. “Virtual Agent Modelling of an Agile Supply Chain Infrastructure”, Management Decision, Vol. 41, No. 7, 2003, pp.625-634. Maes, P. “Agents that Reduce Work and Information Overload”, Association for Computing Machinery. Communications of the ACM, Vol. 37 No. 7, 1994, pp.31-40. Mc Naughton R.B., Osborne P. and Imrie B.C. “Market-oriented Value Creation in Service Firms”, European Journal of Marketing, Vol. 36, 2002, pp.990-1002. Pihkala T., Varamaki E. and Vesalainen J. “Virtual Organisation and the SMEs: A Review and Model Development”, Entrepreneurship and Regional Development, Vol. 11, 1999, pp.335-349. Wang, S. “Meta-management of Virtual Organisations: Toward Information Technology Support”, Internet Research: Electronic Networking Applications and Policy, Vol. 10, No. 5, 2000 pp.451-458. Woodruff, B.W. “Customer Value: The Next Source for Competitive Advantage”, Journal of Academy of Marketing Science, Vol. 25, 1997, pp.139-153. Zadeh, L.: Fuzzy sets.: Inf. Control, Vol.8, 1965; pp.338-353. Zeithaml VA. “Consumer Perceptions of Price, Quality, and Value: A Means-End Model and Synthesis of Evidence”, The Journal of Marketing, Vol. 52, 1988, pp.2-22. Zineldin, M.A. “Towards an Ecological Collaborative Relationship Management. A “co-opetive” Perspective”, European Journal of Marketing, Vol. 32, 1998, pp.11-38.
1630
