Software Agents in E-Commerce Systems

3520

Category: Electronic Commerce

Software Agents in E-Commerce Systems Juergen Seitz University of Cooperative Education Heidenheim, Germany

IntroductIon The Internet introduces a new global marketplace for a large number of relatively unknown and not seldom small companies often offering substitutive or complimentary products and services. The merchants profit from reduced costs, reduced time, and unsold stocks. Customers are attracted by increasing convenience and fast fulfillment. Merchants offering these products and services on this new marketplace need to acquire new customers and sustain ongoing relationships. Nowadays, most merchants’ sites are passive catalogs of products and prices with mechanisms for orders (Dasgupta, Narasimhan, Moser, & Melliar-Smith, 1998). The pull strategy is also applied in auctions available over the Internet, where the seller waits passively for bids. The new push technologies for electronic commerce, like software agents, enable customers to compare a bewildering array of products efficiently, effectively, and automatically (Jennings, Sycara, & Wooldridge, 1998). Switching costs for customers and, thereby, their loyalty to previous suppliers in the marketplace decline (Phlips, 1989; Schwartz, 1999). Using the Internet, the producers profit from reduced cost through direct, non-intermediated sales. The key elements to successful long-term relationships between merchants and customers will be the offering of personalized and value-added services, like one-to-one marketing services, discounts, guarantees, and savings coupons (Seitz, Stickel, & Woda, 2003). In this article, we will analyze possible consequences of new push and pull technologies in electronic commerce for customer’s loyalty, as well as the active technologies enabling customers to purchase efficiently and for the merchants to offer high personalized, value-added, and complimentary products and services. We will discuss some examples of such services and personalization techniques sustaining one-to-one relationships with customers and other actors involved.

Background The World Wide Web provides a great opportunity to compare products and services. Customers as well as competitors may quickly gain detailed and up-to-date data. Especially, suppliers of digital goods are in fear of declining customer’s loyalty. Customers compare catalogs of products of merchants

and producers, and conduct transactions independently of their geographic localization. The crucial basic factors responsible for a limited loyalty of customers are convenience, time, and cost of fulfillment. So, an electronic commerce system should support the ability to embed intelligence to automate the decision process (Dasgupta et al., 1998). The system should not only compare products and prices, but also negotiate and finally purchase products (Teuteberg & Kurbel, 2002). Nowadays, most systems still involve a substantial human element, that is, from the consumer’s perspective neither convenient nor efficient. The human involvement should be limited to transaction specification at the beginning of the process and to the buying or refusal decision at the end of the process (Chen, 2000). This means that an appropriate technology is necessary in the intermediate stages to coordinate between customers and suppliers (d’Inverno & Luck, 2003). Mobile software agents emerge as ideal mediators in electronic commerce and thereby as an appropriate technology for an automated procurement process. Customers may specify constraints on the features of products that enable mobile agents to select products from the merchants’ catalogs and finally to determine the terms of the transaction. Otherwise, software agents may be used by suppliers as market surveyors to determine the current demand and an appropriate price for the good (Chernev, 2003; Fay, 2004; Hann & Terwiesch, 2003; Spann, Klein, Makhlouf, & Bernhardt, 2005; Spann, Skiera, & Schaefers, 2004; Spann & Tellis, 2006). Software agent technology also abolishes the problem of different technological standards, like hardware platforms and operating systems of remote computers. This means that geographical or technological barriers for customers are of no significant importance any more. The key factors are convenience, time, and cost of the procurement process.

agEnt-MEdIatEd ELEctronIc coMMErcE Software agents are computer programs showing the following characteristics (Joshi & Ramesh, 1998): • •

Reactivity: Agent perceives and reacts to environmental changes. Autonomy: Agent has its own program code, data, and execution state.

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

Software Agents in E-Commerce Systems

Figure 1. Classification of software agents (Joshi & Ramesh, 1998)

S

Interaction Multi-user multi-agent Single-user multi-agent Single-user single agent Stationary

Competitive

Collaborative

Cooperation

Highly mobile

Mobility

•

Proactivity: Agent initiates changes to the environment.

The ability of an agent to travel around in networks enhances it to a mobile agent (Brenner, Zarnekow, & Wittig, 1998). Mobile software agents may be classified based on their attributes, like mobility, type of cooperation, and level of interactivity (see Figure 1) (Joshi & Ramesh, 1998). For further possible classifications see, for example, Nwana (1996), Sycara, Decker, Pannu, Williamson, and Zeng (1996), or Kurbel and Loutchko (2001). Competitive agents, mostly single-agents, maximize the interests of their owners. Collaborative agents, on the contrary, share their knowledge and try to maximize benefits of the community on the whole (Joshi & Ramesh, 1998). Mobile agents differ also in terms of the ease of the mobility between remote computers. A continuously traveling nomadic agent, like mobile sales agents (containing information of the total quantity of the product to be sold, the initial price of the product, and the list of potential customers to visit), arrives at a customer’s site and communicates with a stationary customer agent that determines the quantity to be purchased at a given price (Dasgupta et al., 1998). The customer agent uses market values and demand curves of the product for its decision. The sales agent has to adjust the price dynamically during negotiations in order to maximize the gross returns. The price for the product may not be settled too low (an agent sells all of his stock at a bargain price) or too high (a given quantity of the goods may be unsold). Such a supplier-driven electronic commerce system enables merchants to maximize their gross return, but also to identify quickly the customers’ needs and finally to cultivate long-term relationships with them. The architecture of the supplier-driven system was presented by Dasgupta et al. (1998).

From a customer’s perspective, software agents should be highly personalized, continuously running, and autonomous mediators that have to delegate some process management tasks (Guttman, Moukas, & Maes, 1998). A software agent should identify a customer’s needs at first, then retrieve information about the product from the merchants’ sites, compare the offers, and finally determine the terms of the transaction (Castro-Schez, Jennings, Luo, & Shadbolt, 2004). Nowadays, customer agents are mostly used for product and merchant brokerage and for negotiation (Guttman et al., 1998). The price of a product may also be dynamically negotiated instead of being fixed (Phlips, 1989; Schwartz, 1999). For example, tête-à-tête agents cooperatively negotiate multiple terms of a transaction, like warranties, return policies, delivery times, and loan options (Guttman et al., 1998). The buyer agent in a tête-à-tête system negotiates toward a pareto-optimal deal with the sales agent (Fatima, Wooldridge, & Jennings, 2004). A system like this does not maximize gross returns to suppliers or price discounts for customers (Excelente-Toledo & Jennings, 2004; Rahwan et al., 2004). However, it takes into consideration the important value-added merchants’ services (Weerakkody, Currie, & Ekanayaka, 2003). Summarizing, software agents are helping customers to compare and to purchase goods in the Internet. Most of them are agents for a simple online product price comparison or for competitive negotiation over price without considering the value-added and post purchase services from merchants. Such agents decrease customers’ loyalty to a merchant toward zero. However, additional services, like guarantees, return policies, loans, gifts, discounts, and insurance are of interest to customers. Therefore, they should rather use agents comparing or negotiation over multiple terms of a transaction (tête-à-tête). Otherwise, merchants may also send their 3521

Software Agents in E-Commerce Systems

own sales agents to potential buyers in order to acquire new customers and remind the previous customers of new sales offerings (Dutta, Moreau, & Jennings, 2003).

FuturE trEndS In general, software agents helping customers in the procurement process may minimize their loyalty to merchants. Suppliers who do not want to compete solely on the basis of price provide their customers with highly personalized and value-added services to sustain a long-term relationship.

Personalization and Privacy Personalization is defined as the customization of a Web site to meet the particular needs of individual users (Chaffey, Mayer, & Johnston, 2000; Dean, 1998). The goals of personalization technology are to encourage repeated visits and to enhance user loyalty. The identification of customers’ needs occurs through the observation of their behavior and the collection of data (filling out forms or following decision-tree sets of questions). There exist some advanced techniques supporting personalization Web site contents, like rule-based matching and collaborative filtering. Using rule-base matching, users have to answer a set of yes/no or multiple choice questions to settle a set of user’s criteria. Collaborative filtering methods combine the personal preferences of a user with preferences of like-minded people (Dean, 1998). In regard to personalization techniques, one-to-one-marketing should be noticed. This strategy enables targeting unique offers to specific customers (Chaffey et al., 2000). Institutions offering such individualized services have to dispose of accurate user profiles before. A critical factor of personalization is the privacy issue. Filtering and customization techniques entail the collection and the use of personal data, like name, e-mail address, postal address, age, gender, income, Internet connection, and employment status, that must be protected from abuse (Dean, 1998). Furthermore, a lot of suppliers in the Internet deriving revenues mainly from advertising need to identify their users in order to better customize the content and to attract the advertisers being interested. Hence, the user should be informed by suppliers how they use the personal data and how they protect it. Nowadays, there are several initiatives and standardization projects for the privacy of personal data usage. Such initiatives increase user trust and confidence in electronic commerce. However, no organization or institution has the power to enforce it to the wide usage of suppliers.

3522

Value-added Services Merchants who do not want to compete solely on the basis of prices often offer their customers value-added or complimentary services. Complimentary products imply higher benefits to the customer in the case he or she only buys the product he or she looks for (Seitz, Stickel, & Woda, 1999). Such products or services increase the value of the primary good to the customer. Examples of value-added services in electronic commerce are sales discounts, savings coupons, additional insurance and guarantees, gifts, and also free software to test. In general, value-added services enable customers to trade at favorable terms and with confidence. They increase the attractiveness of the merchant to present customers and attract new customers.

reduced Financial transaction costs Internet merchants might achieve additional reduction of transaction costs using electronic payment systems (Bakos, 1998). Electronic payment systems reduce cash handling costs for merchants and improve speed and convenience for customers. The aggregated cost of each payment consists of transformation costs, for example, the fees for conversion from assets to cash and vice versa, transport and storage costs, costs for safety measures, and search and time costs (Hakenberg, 1996).

concLuSIon This article discusses consequences of electronic commerce on customer’s loyalty. Electronic commerce in the Internet offers the possibility to create a perfect marketplace. The intermediation in distribution will be reduced. This means lower costs for both suppliers and customers. Software agents may be classified into different types with regard to their use on supporting electronic commerce. Most of the software agents only perform simple product price comparisons; some support the purchase of products. These software agents reduce customers’ loyalty because the price is the only parameter. Quality and added values are not considered. Therefore, multi-agent systems allowing negotiation might be useful from a customer’s perspective. Merchants may also send their own sales agents to potential buyers in order to remind previous customers of new sales offerings or to suppliers in order to maximize their gross return. Personalization and customization of Web sites, valueadded services, and the reduction of transaction costs are instruments for increasing customers’ loyalty. Personalization techniques, like rule-based matching and collaborative filter-

Category: Electronic Commerce

ing, provide contents on Web sites that are appropriate to the customers’ preferences or analyze past purchases and prior suggestions of other customers. One-to-one-marketing may be especially useful for sophisticated products demanding explanations or to enable cross selling of other products and services. User profiles allow merchants to make customeroriented offers or build special offers including additional services. Value-added services attract the customer to trade at favorable terms. The usage of electronic payment systems may reduce transactions costs.

3523

Excelente-Toledo, C. B., & Jennings, N. R. (2004). The dynamic selection of coordination mechanisms. Journal of Autonomous Agents and Multi-­Agent Systems, 9(1-2), 55-85. Fatima, S., Wooldridge, M., & Jennings, N. R. (2004). Bargaining with incomplete information. Annals of Mathematics and Artificial Intelligence, 44(3), 207-232. Fay, S. (2004). Partial repeat bidding in the name-your-ownprice channel. Marketing Science, 23(3), 407-418.

rEFErEncES

Guttman, R. H., Moukas, A. G., & Maes, P. (1998). Agents as mediators in electronic commerce. EM-­Electronic Markets, 8(1), 22-27.

Bakos, Y. (1998). The emerging role of electronic marketplaces on the Internet. Communications of the ACM, 41(8), 35-42.

Hakenberg, T. (1996). Elektronische Zahlungssysteme im Wettstreit mit dem Bargeld. Sparkasse, 59(6), 271-274.

Brenner, W., Zarnekow, R., & Wittig, H. (1998). Intelligente Softwareagenten. Grundlagen und Anwendungen. Berlin: Springer. Castro-Schez, J. J., Jennings, N. R., Luo, X., & Shadbolt, N. (2004). Acquiring domain knowledge for negotiating agents: A case study. International Journal of Human Computer Studies, 61(1), 3-31. Chaffey, D., Mayer, R., & Johnston, K. (2000). Internet marketing. London: Prentice Hall. Chen, Z. (2000). Intelligent agents. In M. Zeleny (Ed.), The IEBM handbook of information technology in business (pp. 561-569). London: Thomson Learning. Chernev, A. (2003). Reverse pricing and online price elicitation strategies in consumer choice. Journal of Consumer Psychology, 13(1/2), 51-62. Dasgupta, P., Narasimhan, N., Moser, L. E., & Melliar-Smith, P. M. (1998). A supplier-driven electronic marketplace using mobile agents. In Proceedings of the First International Conference on Telecommunications and E-­Commerce (pp. 42-50). Nashville, TN. Dean, R. (1998, June 2). Personalizing your Web site. Retrieved August 29, 2003, from http://builder.cnet.com/webbuilding/pages/Business/Personal/ d’Inverno, M., & Luck, M. (2003). Understanding agent systems. Berlin: Springer. Dutta, P. S., Moreau, L., & Jennings, N. R. (2003). Finding interaction partners using cognition-based decision strategies. In Proceedings of the IJCAI workshop on Cognitive Modeling of Agents and Multi-­Agent Interactions (pp. 4655). Acapulco, Mexico.

Hann, I.-H., & Terwiesch, C. (2003). Measuring the frictional costs of online transactions: The case of a name-your-ownprice channel. Management Science, 49(1), 1563-1579. Jennings, N. R., Sycara, K., & Wooldridge, M. (1998). A road map of agent research and development. Autonomous Agents and Multi-­Agent Systems, 1(1), 7-38. Joshi, N., & Ramesh, V. C. (1998). On mobile agent architectures (Tech. Rep.). Illinois Institute of Technology, ECE Department, Chicago, IL. Loutchko, I., & Kurbel, K. (2001). A framework for multiagent electronic marketplaces: Analysis and classification of existing systems. In Proceedings of International ICSC Congress on Information Science Innovations (ISI 2001), American University in Dubai, U. A. E. (pp. 335-339). Nwana, H. S. (1996). Software agents: An overview. The Knowledge Engineering Review, 11(3), 205-244. Phlips, L. (1989). The economics of price discrimination. Cambridge: Cambridge University Press. Rahwan, I., Ramchurn, S. D., Jennings, N. R., McBurney, P., Parsons, S. & Sonenberg, L. (2003). Argumentationbased negotiation. The Knowledge Engineering Review, 18(4), 343-375. Schwartz, E. I. (1999). Digital Darwinism: 7 breakthrough business strategies for surviving in the cutthroat Web economy. New York: Broadway. Seitz, J., Stickel, E., & Woda, K. (1999). Electronic payment systems: A game-theoretic analysis. In M. KhosrowPour (Ed.), Managing information technology resources in organizations in the next millennium. Proceedings of the 1999 Information Resources Management Association International Conference (pp. 564-568). Hershey, PA: Idea Group Publishing.

S

3524

Category: Electronic Commerce

kEy tErMS

Seitz, J., Stickel, E., & Woda, K. (2002). Impacts of software agents in e-commerce systems on customer’s loyalty and on behavior of potential customers. In B. Fazlollahi (Ed.), Strategies for e-­commerce success (pp. 208-223). Hershey, PA: IRM Press.

Collaborative Filtering: Collaborative filtering methods combine personal preferences of a user with preferences of like-minded people to guide the user.

Spann, M., Klein, J., Makhlouf, K., & Bernhardt, M. (2005). Interaktive Preismassnahmen bei Low-Cost-Fluglinien. Zeitschrift fuer Betriebswirtschaft (ZfB), 75(EH1), 53-77.

Customer Loyalty: Because there is no existing ownership to service products, suppliers have to make special efforts to get long-standing customers.

Spann, M., Skiera, B., & Schaefers, B. (2004). Measuring individual frictional costs and willingness-to-pay via name-your-own-price mechanisms. Journal of Interactive Marketing, 18(4), 22-36.

Customer Profiling: Usage of the Web site to get information about the specific interests and characteristics of a customer.

Spann, M., & Tellis, G. (2006). Does the Internet promote better consumer decision? The case of name-your-own-price auctions. Journal of Marketing, 70(1), 65-78. Sycara, K., Decker, K., Pannu, A., Williamson, M., & Zeng, D. (1996, December). Distributed intelligent agents. IEEE Expert. Retrieved July 11, 2006, from http://www.cs.cmu. edu/~softagents/papers/ieee-agents96.pdf Teuteberg, F., & Kurbel, K. (2002). Anticipating agents’ negotiation strategies in an e-marketplace using belief models. In W. Abramowicz (Ed.), Proceedings of the Fifth International Conference on Business Information Systems (BIS 2002) (pp. 91-100). Poznan, Poland: Akademia Ekonomiczna w Poznaniu, Katedra Informatyki Ekonomicznej. Weerakkody, V., Currie, W. L., & Ekanayaka, Y. (2003). Reengineering business processes through application service providers—Challenges, issues and complexities. Business Process Management Journal, 9(6), 776-794.

Customization: The adjustment of products or services to individual needs. Basic characteristics are implemented in the product or service and may be controlled by parameters. Disintermediation: The elimination of agents, like wholesale dealers or brokers, who built the former relationship between producer and consumer. Disintermediation allows the direct supply of the consumer. One-to-One-Marketing: The direct dialog between a producer and an individual consumer or a group of consumers with similar needs.S Personalization: Web-based personalization means providing customized content to individual users using Web sites, e-mails, and push technologies. Software Agents: Computer programs that are characterized by reactivity, autonomy, and proactivity. Therefore, the software agent interacts with its environment.