This paper looks at how customer-to-customer (C-to C) interaction on computer networks is becoming increasingly important and explores the dynamics involved. The message is quite simple. In the age of computer networks, sellers of products should be aware that their customers communicate with each other much more extensively than before. The reputation of a product, for example, is disseminated much more quickly and extensively than ever before. Recently one finds numerous entirely user-organized home pages of almost all kinds of product on the Internet. To give another example, one regularly encounters a situation in which users of a product help each other on a computer network. Thus C-to-C interaction extends to voluntary product support.

All of these factors change the nature of the market significantly and require firms to change their strategies in facing customers. This paper explores the nature of the cyberspace economy through the analysis of C-to-C interaction on computer networks. Pieces of qualitative and quantitative evidence are used for the analysis.

C-to-C interaction on computer networks is defined as “the interaction among customers (of a certain product or service) on computer networks that affects the purchase, the value, and/or the conceived satisfaction toward the product. Theoretically, C-to-C interaction can be described as interaction among agents in polyagent systems (Takagi et. al (1995)) that are networked by computer networks and standardized interface.

The importance of interactiveness in management has been discussed by many writers including Norman and Ramirez (1994),and Yahagi et. al (1996). The message is straightforward. In affluent societies, it is not sufficient for firms to produce fine products. Firms must listen to and incorporate customer needs to their products and product delivery systems interactively. The emphasis in marketing on “relationships” (Shimaguchi (1995)) reflects such a view.

Much of the discussion surrounding interactiveness, however, centers on the interactiveness between the sellers of products and their customers. We call this seller-to-customer S-to-C interaction. While admitting the importance of S-to-C interaction, the primary message of this paper is that the relationship among customers themselves, which we call C-to-C interaction, is becoming even more important.

The penetration of more “open” computer networks is the context in which C-to-C is flourishing. Today’s computer networks, in contrast with the traditional “closed” system, allow tens of millions of computers all over the world to communicate freely at low cost. This environment allows consumers -who were previously secluded from each other- to exchange large quantities of information. They share information, feed it back to the producer, and even participate in productive activities.

C-to-C interaction is probably already a daily practice for those readers that participate in electronic forums. When you encounter a problem with your personal computer, it is usually quicker and more pleasant to ask your fellow networkers than making a telephone call to customer support personnel. It is amazing how total strangers are willing to spend a lot of time for each other for no monetary reward.

While it is easy to predict that C-to-C interaction will grow and that it will have a significant impact on the operations of firms, it is not easy to predict how C-to-C interaction will actually influence corporate operations. We know very little about the nature and the dynamics involved. Even tougher is the task of advising firms how to deal with it. Our research, which will be presented later, indicates that the attempt of firm to manipulate the C-to-C interaction process can actually suppress it.

1.1 Interactiveness

Before discussing C-to-C interaction further, it would be helpful to map various types of interaction in the marketplace. Figure 1 is a simplified representation of the various types of communication between sellers of products and their customers.

“One-way communication between the seller and the customers” shows the traditional situation in which the seller provides the product feature and availability information to the customers with no feedback from them. The conventional mass marketing communication channels such as television advertisements, served this purpose well. When there is excess demand in the market, one-way communication suffices.

When the market environment becomes more challenging, firms need to establish two- way communication with the customers to reflect their views in the product development and delivery. Information technology offers powerful tools for firms that attempt to “mass customize” (Pine(1993)) to customer needs. At the core is the database in which profiles of each customer are accumulated. In order to acquire the customer data, many related technologies, such as the magnetic card (and IC smart card) reader and bar code, are employed. More recently, world wide web (WWW) home pages are being designed so that visitors’ profiles are automatically accumulated in the hosts computers.

At the theoretical level, two-way S-to-C interaction implies an emphasis on “relationship management,” as advocated by Shimaguchi (1995). Such management style assumes the existence of two-way communication channels.

The third pattern of communication is C-to-C interaction, which is the subject of this paper. This interaction goes beyond the traditional S-to-C interaction i.e., customers begin to communicate with each other. They exchange product assessment, availability information, and give advice to each other on the purchase and the use of products.

In electronic conference rooms, consumers offer information on new products, offer help to other users, and even develop tools for other users. Most often, they do so not for profit incentives but on a voluntary basis (Kaneko (1994)). Such voluntary information, because it is not offered with commercial incentives, is often more credible than the commercial information offered by the sellers. Thus it is sometimes very influential in the sales of the products. We can often identify opinion leaders in electronic forums that influence the user community strongly. The endorsement of products by such opinion leaders in cyberspace is becoming increasingly important even for the sellers of “real products” in real space.

1.2 Platforms of C-to-C Interaction

It is important to note here that C-to-C interaction does not happen automatically. As noted in Figure 1, C-to-C interaction requires platforms of communication that fulfill the premises for successful communication among partners. Here we identify two such premises.

First is the physical existence of communication channels. When customers that are potential interaction partners are physically apart and lack communication means, C-to-C interaction cannot materialize.

When the products involved are related to our daily lives, we observe C-to-C communication occurring face-to-face without electronic channels. Although the scene is becoming rarer now in Japanese society, it was part of our lives to see housewives on street corners exchanging information on the best buy groceries of the day. This face-to-face communications, however, requires the communication partners to share time and space, which is becoming increasingly difficult in our lives today.

Computer networks, on the other hand, allow customers who were traditionally secluded by distance to overcome the obstacle. In addition to its power in overcoming distance, computer-mediated communication has such functional features as (1) asynchronous communications, (2) multipoint communications, and (3) interactive communications. These allow combinations of individuals who were previously isolated from each other to communicate. These added capabilities allow forums of broader topics to sustain themselves.

Second, C-to-C communication requires the sharing of language and culture among the customers. The mere existence of physical channels of communication is not sufficient for communication partners to exchange messages in a meaningful way. Kokuryo (1996) identified sharing of the following as the necessary conditions for successful communication between partners: (a) symbols (vocabulary), (b) grammar, (c) scenarios, and (d) norms.

In the latter part of this paper, we will see the case of Internet Go Server. Go is a traditional board game on which two (or two pairs) of players compete. Much like the game of chess, this game has clearly defined rules, a set of vocabulary that is shared globally, and accumulation of shared experience (in the form of records and tales of historic games). The existence of such common language allows players around the world who do not share spoken language to play very smoothly. The existence of such low barriers to entry and the open atmosphere that the forum creates are attracting newcomers to enter the world.

The question of whose language becomes the global standard of C-to-C communication is clearly a very sensitive issue. English is evidently becoming a common language. Most of the communication on IGS, for example, is conducted in English, although the game itself is clearly Asian.

2 Cases Studies of C-to-C Interaction

Prior to a quantitative study, the research team has been conducting case studies on several instances of C-to-C interaction to identify its nature. It would be helpful at this stage to briefly introduce a few illustrative cases of C-to-C interaction . They served as the basis for the development of analytical models and the design of the empirical research. Here we present three cases.

Amidst growing enthusiasm, fans of the HP machine began to organize themselves in an electronic conference room. They divided the tasks of developing necessary software and fonts among themselves and uploaded the output in the electronic library for anyone to use for free.

An interesting observation in this research has been the existence of “real place” and a “focal person” in the process. The real place was a store in Shinjuku in Tokyo, which is one of the largest shopping districts in Japan. The office goods store that carried the HP products was known among enthusiasts as the Shinjuku Kyokai (Church). The dealer had a sales engineer who clearly played the focal role in C-to-C interaction. The users would flock to the store to see the engineer with the machine, exchange their findings and would go home to log on to the electronic conference room, where a much larger body of enthusiasts awaited.

2.2 Internet Go Server (IGS)

Internet Go Server is a server on the internet that offers playing ground for the players of “go” WHICH IS a traditional Asian game (Takegawa (1995)). The game has relatively simple structure with two people(or sometimes two couples) taking turns to place black and white stones on a 19 X 19 board to compete for the enclosure of the larger territory. While the rules are very simple, the game can be very deep and there are professional go players around the world and many prized competitions.

The server was originally developed by go fans in the United States and was placed in American universities. As of August 1996, the server had over 10,000 members globally, with international matches being held regularly. Both the server software and the client software were developed by go fans. Interface to the server is disclosed to which user developed client software.

As the number of clients increased, it became increasingly difficult to maintain the server on a voluntary basis. Thus in 1995, the rights to the server were sold to a commercial Internet service provider in Korea. A company in Japan that has an alliance with the Korean company is trying to offer the service on a commercial basis. However,it is struggling to win user acceptance for the commercial offering of a once volunteer- based free service.

2.3 JW-CAD

JW-CAD is a computer-aided design software for personal computers that was offered for free on a computer network (Ayabe (1995)). At the time of its launch, the performance excelled some of commercial software packages that were sold for thousands of dollars. This gave the software industry quite a shock. Since it was offered in 1991, the software enjoyed strong user support and won the Nikkei Computer Graphics best CAD award for two consecutive years.

A notable characteristics of “freeware,” such as JW-CAD, is the existence of strong user support among themselves in various electronic forums. Instead of commercial software packages, to which customers expect full support from the vendors, users of freeware often consider the product their own and are willing to help other users and the developers of the product.

It is interesting to observe that similar phenomena occur for some “shareware,” which is copy-free software for which the developer charges a small fee. This shows that when customers of a product sympathize with a product, they are willing to help the commercial sellers. It is evident that it is the strong supporters that are allowing Macintosh computers to sustain a certain level of market share in Japan in spite of their difficulties.

In the case of JW-CAD, when it was attacked by commercial software developers, who claimed that JW-CAD had questionable reliability and a lack of formal product support, strong voice (in the form of e-mails) was raised by the user group in support of the freeware. The hostile actions by commercial developers subsequently subsided.

3. Typology of C-to-C Interaction

The analyses of several cases including the above three, reveal that there are in fact several types of C-to-C interaction, which seem to share a similar mechanisms but are different in their roles in the market. It is therefore useful to categorize and analyze them individually.

3.1 Category 1: Reputation (“Word of E-mail”)

C-to-C interaction on a computer network is becoming a key source of product information for consumers. Instead of acquiring product information from the manufacturers or their dealers, today’s consumers rely on their peers in cyberspace to give them honest and up-to-date information. In other words, computer networks provide very large and efficient mechanism for word-of-mouth (we might call it “word of e-mail”).

Itami (1984) identified two ways in which “customer interactions” affects the performance of firms as follows: (a) the support of a customer for a product gives invisible asset to the firm in the form of product credibility, brand recognition and word-of-mouth advertising channels, (b) the demonstration effect, in which a crowded sushi-bar attracts even more customers. In this section we combine these and call them the “formation of reputation on electronic communication channels.”

Word-of-mouth and reputation are nothing new, they have been in existence ever since businesses started. At the same time, computer networks give a whole new meaning to this very old practice by enlarging the geographic coverage and the scope of activities to a global level.

The author is one of the many who received the first news of a bug in famous manufacturer’s processor a few years ago through an electronic mail from a friend. The news was e-mailed to me from several of my friends in a very short period of time before it started to appear in electronic bulletin boards. It was a few days after that that newspapers started on write about it. The official announcement lagged even further.

To give another example, in 1995, when the Tokyo subway was terrorized by a poison gas attack, it was apparent that various types of rumor was being circulated in cyberspace at an enormous speed and were regenerating themselves. In one instance, such a rumor resulted in the shutdown of a Shinjuku shopping facility, causing severe economic and social damage.

News both good and bad circulates in cyberspace at an unprecedented speed. On the network, a species of people lives that loves to test products as soon as they are launched and break news of what is good and bad about them. In many cases they have charismatic influence and affect the sales of the product.

3.2 Category 2: User Support

This refers to the phenomenon of user groups of a product helping each other in trouble- shooting. Such mutual support is seen most often when a customer of an electronic product, such as a PC, purchases a product and encounters problems.

The writer is probably one of the great many people who has had unpleasant experience in making telephone calls to manufacturer support centers to ask questions. To begin with, they do not answer the phone. Even when they do, they often fail to understand the customer’s circumstances. Users who call the support centers are often bewildered by the technical jargon that the customer-support person uses. On the other side of the phone, customer-support people are often frustrated by customers’ inability to express their problems structurally.

JW-CAD is a powerful example of how C-to-C user support played a critical role in the survival of a software package. JW-CAD had an electronic user group conference room in a consumer computer service, Nifty-serve. The existence of the user group support more than compensated the weak product support of a freeware. In the conference room, questions of novices are quickly answered by other more experienced users who have experienced the same situations. This was a major threat to the commercial software developers that had to hire many personnel for product support. As discussed in the case of JW-CAD, such user participation in the support of a product occurs not only for freeware but also for commercial products.

Regardless of whether the commercial product suppliers are aware of it or not, C-to-C product support is creating value even for them. If every single question that consumers have on about an automobile came to the manufacturers or the dealers, how much customer support would they need? Such C-to-C product support existed for even before computer networks, but now it is expanding rapidly in the emerging electronic environment.

3.3 Category 3: Product Development

C-to-C interaction often goes beyond the user support level and into the development of supplementary products (such as plug-ins, fonts, and utility software) by the customers themselves. The existence of a large library of free supplemental software adds great value to the core product. IGS started as a free service organized by a volunteer group. The interface to communicate with the server was disclosed to the user community, thereby allowing the development of multiple client software.

To give another example, JW-CAD created an opportunity for an unrelated third- party company to publish and sell a JW-CAD manual with a floppy disk of the software attached. Users of CAD could simply buy the manual with the floppy from a bookstore for a little over US$40 and acquire full CAD capability, which would cost thousand of dollars with commercial software at the time.

These examples indicate that “customers” in C-to-C interaction can be commercial customers that provide commercial services on the platform. To explain such phenomena, Imai and Kokuryo (1994) proposed the concept of “platform business,” which is defined as follows: Businesses that provide, as a private entity and in an open manner (i.e., with explicit and fair conditions to anyone who wishes to receive them), such products or services that stimulate trade among third parties and/or the creation of new businesses.

Platform businesses provide an environment in which customers to the business create businesses to which other members of the platform become customers (Kokuryo and Takeda (1995)). The relationship between telephone companies and the Internet service providers can be portrayed as a dramatic example of commercial C-to-C interaction on platform businesses. Most of the service providers on the Internet are in fact customers of the telephone companies’ switched- and leased-line services. In the old days, users of telecommunication services had to wait for the telephone companies to start new services. Now the users of the Internet wait for other users of Internet to come up with the next block buster application. In the meantime, the application of the telephone companies lines expands both quantitatively and functionally. Thus the network “evolves,” as Aizu (1994) put it.

3.4 Category 4: C-to-C Interaction as a Goal in Itself

C-to-C interaction may be the service itself in many instances. IGS, for example, offers an environment in which members play with each other for the sake of it. When a member logs on to the system, the list of members presently on line is exhibited with information on their identity and the level of playing skill. The system thus matches players and provides functions to play games.

At the conceptual level, C-to-C interaction is the biggest added value that providers of telecommunications networks, such as telephone companies and Internet providers, offer. What good is a telephone network if there is no one on the other side to answer your phone?

One important role of a network provider is in the determination of common interface. They provide various levels of standardized interface that customers adopt to exchange information with each other. Without such standards, effective communications do not materialize.

3.5 Category 5: The Bandwagon Effect

While somewhat different from the other four categories, we should probably include the formation of critical mass as the fifth category of C-to-C interaction. Usually called the bandwagon effect, the adoption of a product by a consumer is affected by the adoption of others.

Network externality is the reason why the bandwagon effect is particularly relevant in the network environment. Network externality refers to the phenomenon of the utility of a product increasing as the number of adopters to it increases. The utility of network services such as the telephone, increases as the number of subscribers increases. The attractiveness of platforms, such as operating software for computers, increases as the number of their adopters increase. It seems that the bandwagon effect is becoming ever more important as various media, including computer networks, integrate previously fragmented markets.

4. Empirical Research

4.1 Outline of Research

Based on recognition of the growing role of C-to-C interaction, a team at Keio Business School began an empirical research on the nature of C-to-C interaction on computer networks. In 1996 Tamura, in his master’s thesis project, conducted research on the conditions that stimulate C-to-C interaction (Tamura (1997)).

More specifically, we looked into sampled electronic forums in Nifty-serve, which is a major commercial electronic services company. Nifty-serve started as an adaptation of CompuServe services to the Japanese environment and has a structure similar to it. There are many electronic conference rooms on the Nifty-serve platform that are organized around one or more product of a company. Many users of the products actively communicate with each other in the environment.

Randomly sampled electronic conference rooms are analyzed to test hypotheses on the instances in which C-to-C interaction become active. We have used the messages that users exchange among each other as the index of C-to-C interaction. The research was exploratory, in that the researchers allowed the model and hypotheses to evolve in two stages.

4.1 Stage One: A Comparison of Corporate-Sponsored “Stations” versus Users’ Voluntary “Forums”

As the first stage in our research, we compared “forums” and “stations” within Nifty-serve that are organized around a particular product or group of products. They both have electronic conference rooms in which users exchange messages and libraries in which users share software that are offered either free or for money as shareware.

While almost identical in functional aspects, forums and stations differ in an important aspect. Specifically, stations are corporate sponsored by the manufacturers of the products, while forums are established as users’ voluntary communication space.

Reflecting the nature of the ownership structure, the participants in the electronic conference rooms are subtly different. Members to stations include (1) manufacturers’ representatives, (2) “heavy users,” defined in this research as participants in the electronic conference rooms who frequently participate in discussions and give advice to other users on the use of the products, (3) ordinary users who participate in the exchange of messages in the electronic conference rooms but do not give advice to others, and (4) ROM (read-only members), who read the messages but remain totally silent. In comparison, forum members include the heavy users, the ordinary users, and the ROMs, but not the manufacturers representatives.

The comparison between the forums and the stations was considered closely related to the ultimate motivation of the research, which was to provide advice to firms on how to deal with C-to-C interaction. At the same time, the research team anticipated that C-to-C interaction would be more active in forums in which there is an atmosphere of users forming a community of their own, as opposed to stations, in which users expects manufacturers to provide them with the service. Thus we decided to test the following hypothesis:

Hypothesis 1: The level of C-to-C interaction is higher in voluntary electronic conference rooms (i.e., in forums) than in seller-sponsored ones (i.e., in stations).

To test this hypothesis, the researchers randomly sampled 30 electronic conference rooms each from stations and forums. The researchers started by testing the difference of the average number of messages between the forum and the station. Table 1 indicates the results:

Table１

　　　　　　　　　　　　　　　　　　(Note:Significant at 99% level.)

Hypothesis 1 thus is supported. While the external validity of this result is debatable, corporate ownership of the communications platforms seemed to suppress the level of C-to-C interaction. (See footnote .)

4.2 Stage 2: The Role of Heavy Users and Manufacturer Representatives in Station Electronic Conference Rooms

Based on the confirmation that manufacturer ownership of the communication platform has a negative effect on the level of C-to-C interaction, we designed our second stage analysis that focused on the stations. More specifically, we attempted to identify (a) whether and how the behavioral characteristics of manufacturer representatives might affect the level of C-to-C interaction, and (b) the role of heavy users in stimulating C-to-C interaction.

As for (a),we speculated that C-to-C interaction will be more active when manufacturer representatives behave as a “friend” or an “equal member” of the user community and avoid being the “them, the service providers” among the “us, the users.”

We thus determined the following set of hypotheses:

Hypothesis 2: The attribute of manufacturer representatives emphasizing their role as the product provider suppresses C-to-C interaction, whereas their friendly attitude (which makes them resemble heavy users) stimulates C-to-C interaction.

To make this hypothesis testable, it was broken down to the following:

Hypothesis 2.1: When manufacturer representatives use predetermined “form responses” to many users, we will observe low levels of C-to-C interaction.

Hypothesis 2.2: The large difference between the users and the manufacturer representatives in the use of the phrase ne which is used in the Japanese language among close friends suppresses C-to-C interaction.

Hypothesis 2.3: The efforts of manufacturer representatives to initiate discussions by offering topics for discussion stimulates C-to-C interaction.

Hypothesis 3: The existence of heavy users in stations stimulates C-to-C interaction.

To test these hypotheses, we randomly selected 30 electronic conference rooms from stations that were more than two years old. Then the last 100 messages of each electronic conference rooms observed on October 31, 1996, were used as data. Messages were categorized into (a) C-to-C interaction, (b) S(seller)-to-C interaction, and (c)noninteractive (stand-alone) messages. The number of C-to-C interaction was calculated by subtracting noninteractive messages and S-to-C interaction from the total.

The existence of heavy users in a conference room was determined by identifying a user that (a) sent messages more frequently than other users and (b) provided advice to other users on the use of the product involved. The use of form responses was identified by observing the manufacturer representatives’ practice of using the exactly same phrases to respond to multiple users.

The use of “ne” which is a suffix used in the Japanese language at the end of sentences to show friendliness, was considered an index of whether the manufacturer representatives identified themselves simply as a member of the community or as someone that plays a special role. In this research, the absolute difference in the level of ne use was adopted as the index.

Table 2 and Table 3 show the results of the tests.

Table 2: Test results for hypothese’s group 2 (on manufacturer representatives’ attitudes)

Note : Indicates rejection of the hypothesis at 95% significance level.

Table 3: Test results for hypothesis 3 (existence of heavy users)

4.3 Analyses of Test Results

Table 2 indicates that hypotheses 2.1, 2.2, and 2.3 were not supported. Not only that, the test of hypothesis 2.2 indicated significant results in the reverse direction. Table 3 indicates that hypothesis 3 was supported.

The rejection of hypothesis 2.2 was the most striking result to the researchers. This suggests that the “friendly” attitude of the manufacturer representatives is not only ineffective but can even disrupt for C-to-C interaction.

We can only hypothesize the meaning of this result. It seems that the customers are simply telling the manufacturers to stay out of customers’ voluntary interaction. In other words, the hypocritical intrusion of the manufacturer representatives destroys the atmosphere of self-help.

While it is premature to make any definitive judgment, this result makes us think that it is perhaps impossible for manufacturers to play an active role in C-to-C interaction in a direct way and that we should advise firms to simply stay out. We should at least say that firms should be cautions that mindless intervention in C-to-C interaction can backfire on them.

As opposed to the test on the role of manufacturer representatives’ attitudes, the hypothesis on the role of heavy users was supported at a high significance level. Qualitative examination of the actual interaction that took place in the electronic conference rooms shows that heavy users are not only instrumental in C-to-C interaction but also play an important role in resolving emotional conflict between manufacturers and customers. At the same time, our interview revealed that manufacturer representatives are often afraid of heavy users, who dominate the atmosphere of the user community.

Heavy users are wonderful collaborators as long as they are supportive to the manufacturers. In some cases, however, heavy users often become the leader of rebellions or attacks the manufacture’s. In such cases, they become a formidable foe. Manufacturer representatives are aware of this dynamics,and we often find manufacturers trying to patronize heavy users. Such an attempt may be one way for firms to influence C-to-C interaction. Even this, however, sometimes backfires on them when an attempt to manipulate the C-to-C interaction is revealed.

5. Discussion

It seems evident to the author that C-to-C interaction is becoming increasingly important as the open computer network develops. It makes markets more efficient (as in the case of reputation), as well as creating value (as in product support and product development).

Very little is known, however, about the nature of C-to-C interaction on computer networks. While there is marketing literature on the subject, this was developed before the explosion of the Internet and it is not clear how theories of face-to-face interaction may apply to the phenomena in cyberspace.

Even more difficult is the task of advising firms on how to deal with C-to-C interaction. The research that we have conducted so far suggests that any attempt by commercial firms to manipulate C-to-C interaction not only fails but also invites retaliation for the intervention. Many experts on cyberspace to whom we consulted on the results of this research have agreed with the view. Some suggested that the only pragmatic approach might be to advise firms on C-to-C crisis management i.e., to advise firms how to avoid negative publicity on the net and/or how to cope with crises once they have occurred.

At the same time, we clearly see some companies that are enjoying strong support from friendly C-to-C interaction. Traditionally we would see users of Honda and Sony enthusiastically exchanging their experience with the products, preaching the concept and helping novice users get used to them. Today computer game enthusiasts exhibit similar behavior on some of the game hardware and software.

Such observations indicate that while it may not be possible for firms to directly influence C-to-C interaction, firms should present themselves and their products so that they create strong sympathy among customers. How that can be done will be the topic of our future research.

References: (Translation of Japanese titles to English by the author)

Aizu, Izumi (1994), Shinka suru Network (The Evolving Network), NTT Publishing

Ayabe, Motoko(1995)，JW-CAD，Keio Business School Case Study

Imai, Ken-ichi and Jiro Kokuryo eds. (1994), Platform Business (in Japanese), Infocom Research

Itami, Hiroyuki (1987)、Invisible Assets,Harvard University Press

Kaneko, Ikuyo (1994), Volunteer: Mou Hitotsu no Jouhou Shakai (Volunteering: Another Form of the Information Society), Iwanami

Kokuryo, Jiro (1996), “Keiei Senryaku toshiteno Interface no Open-ka” (The Adoption of Open Interface as a Business Strategy,in Business Insight, Summer 1996

Kokuryo,Jiro and Yoko Takeda (1996) “Role of Intermediaries in Electronic Commerce,” paper submitted to OECD Workshops on the Economics of the Information Society Workshop No. 2, Istanbul (http://www.oecd.org/dsti/iccp/turkagen.html)

Normann, Richard and Rafael Ramirez (1994), Designing Interactive Strategy, John Wiley & Sons

Pine II, B. Joseph(1993), “Mass Customization,” Harvard Business School Press

Shimaguchi, Mitsuteru (1995), Interactive Marketing no Seiritsu Jouken to Kadai (Interactive Marketing -Conditions for its Development and Issues), Keio Business School Research Paper Series No.32.

Takagi, Haruo, Kyoichi Kijima, and Hiroshi Deguchi (1995)， Multimedia Jidai no Ningen to Shakai -Polyagent Society- (Individuals and Societies in the Multimedia Age -A Polyagent Society-，Nikka Giren Publishing

Takegawa, Zenta (1995)，Internet Go Server(in Japanese)，Keio Business School Case Study

Tamura, Takashi(1996)， “Computer Network ni Okeru Kokyaku kan no Sougo Sayou Soushutu Kouka ni Tsuite” (On the Creation of Customer-to-Customer Interaction on Computer Networks), Master’s Thesis submitted to Keio Business School

Yahagi, Tsuneo, M. Aoi, M. Shimaguchi, and M. Wada (1996), Interactive Management, Diamond