Hierarchical Structure of Intranet Functions and Their Relative Importance:: Using the Analytic...

16
Ž . Decision Support Systems 23 1998 59–74 Hierarchical Structure of Intranet Functions and Their Relative Importance: Using the Analytic Hierarchy Process for Virtual Organizations Jinwoo Kim ) Department of Business Administration, Yonsei UniÕersity, Seoul 120-749, South Korea Abstract The Intranet has clear potentials to provide necessary tools for building virtual organizations which are essential in the mass-customization economy. However, the functions to be included in the Intranet system, and their importance in building virtual organizations have not been examined yet. The main objectives of this study are to construct an analytic structure of Intranet functions and to measure the relative importance of the functions. In order to achieve such objectives, a hierarchical structure of Intranet functions was built based on interviews with industry experts. Next, using the Analytic Hierarchy Ž . Process AHP , this study conducted surveys with three different groups of people: top management, middle management, and low management. Results from the surveys indicate that several Intranet functions were evaluated as unanimously more important, while others were perceived with varying degrees of importance depending upon the organizational level. This study concludes with the implications of the results to the development of Intranet systems. q 1998 Published by Elsevier Science B.V. All rights reserved. Keywords: Intranet; Virtual corporation; Analytic Hierarchy Process 1. Introduction The trend of globalization is responsible for the transition of the existing mass production economy to a consumer-oriented mass customization economy w x 13,21,26,27 . The goal of mass customization is to provide virtual products, a novel concept which sig- nifies delivering a product or service to consumers at wx any time, to any place, in any form on demand 4. To accomplish this goal, corporations must shift ) E-mail: [email protected] from the presently rigid hierarchical structure to a new flexible organizational structure called a virtual w x organization 7,8 . A virtual organization can be compared to a baseball all-star team where players who excel in their respective positions gather to form a team and compete on a one-time basis. After the game ends, the players return to their teams and carry out their regular duties until another such occasion calls upon them. Similarly, a virtual organi- zation is a temporary union of companies that pos- sess expertise in a specific field and gather for a wx particular project 7 . After the project ends, the group disbands and the companies return to their daily tasks until a new project arises. 0167-9236r98r$19.00 q 1998 Published by Elsevier Science B.V. All rights reserved. Ž . PII S0167-9236 98 00036-0

Transcript of Hierarchical Structure of Intranet Functions and Their Relative Importance:: Using the Analytic...

Page 1: Hierarchical Structure of Intranet Functions and Their Relative Importance:: Using the Analytic Hierarchy Process for Virtual Organizations

Ž .Decision Support Systems 23 1998 59–74

Hierarchical Structure of Intranet Functions and Their RelativeImportance:

Using the Analytic Hierarchy Process for Virtual Organizations

Jinwoo Kim )

Department of Business Administration, Yonsei UniÕersity, Seoul 120-749, South Korea

Abstract

The Intranet has clear potentials to provide necessary tools for building virtual organizations which are essential in themass-customization economy. However, the functions to be included in the Intranet system, and their importance in buildingvirtual organizations have not been examined yet. The main objectives of this study are to construct an analytic structure ofIntranet functions and to measure the relative importance of the functions. In order to achieve such objectives, a hierarchicalstructure of Intranet functions was built based on interviews with industry experts. Next, using the Analytic Hierarchy

Ž .Process AHP , this study conducted surveys with three different groups of people: top management, middle management,and low management. Results from the surveys indicate that several Intranet functions were evaluated as unanimously moreimportant, while others were perceived with varying degrees of importance depending upon the organizational level. Thisstudy concludes with the implications of the results to the development of Intranet systems. q 1998 Published by ElsevierScience B.V. All rights reserved.

Keywords: Intranet; Virtual corporation; Analytic Hierarchy Process

1. Introduction

The trend of globalization is responsible for thetransition of the existing mass production economyto a consumer-oriented mass customization economyw x13,21,26,27 . The goal of mass customization is toprovide virtual products, a novel concept which sig-nifies delivering a product or service to consumers at

w xany time, to any place, in any form on demand 4 .To accomplish this goal, corporations must shift

) E-mail: [email protected]

from the presently rigid hierarchical structure to anew flexible organizational structure called a virtual

w xorganization 7,8 . A virtual organization can becompared to a baseball all-star team where playerswho excel in their respective positions gather to forma team and compete on a one-time basis. After thegame ends, the players return to their teams andcarry out their regular duties until another suchoccasion calls upon them. Similarly, a virtual organi-zation is a temporary union of companies that pos-sess expertise in a specific field and gather for a

w xparticular project 7 . After the project ends, thegroup disbands and the companies return to theirdaily tasks until a new project arises.

0167-9236r98r$19.00 q 1998 Published by Elsevier Science B.V. All rights reserved.Ž .PII S0167-9236 98 00036-0

Page 2: Hierarchical Structure of Intranet Functions and Their Relative Importance:: Using the Analytic Hierarchy Process for Virtual Organizations

( )J. KimrDecision Support Systems 23 1998 59–7460

There are many requirements in establishing avirtual organization. A virtual organization must havea new vision, a common goal among companies,dependable suppliers, and customers in a trust-based

w xrelation 4 . First and foremost, however, a virtualorganization must have a flexible and seamless infor-mation network among different companies dis-persed around the world. This is because informationwill be the core of a virtual organization, and theorganization’s ability to create products will be de-pendent on its information-gathering, -processing,

w xand -integration skills 8 .Nowadays, the Internet has enabled companies to

form a virtual organization where partners or work-ers located over a wide area are linked seamlesslyw x19 . The emerging concept of using Internet technol-ogy as the information system of a virtual organiza-tion is the Intranet. The Intranet is defined as the useof Internet technologies within an organization toachieve better results than the conventional means of

w xdata access and transfer 1,10,20,22,28 . However,due to the relative recency of the Intranet system, ageneral consensus on the functional structure and therelative importance of its various functions is yet tobe reached. This lack of consensus has resulted inambiguities on which functions are to be included inthe Intranet system and how important they are todevelop virtual organizations. Such ambiguities haveled companies to define the Intranet system accord-ing to their own interpretations. For example,database development companies define the Intranetsystem as the integration between the Internet anddatabase, and groupware development companies de-fine it as the integration between the Internet andgroupware. However, these assertions only touchupon one aspect of the Intranet system, while a morecomprehensive discernment is in dire need in orderto realize its full potential.

The main objectives of this study are to constructan overall structure of the Intranet functions and tomeasure the relative importance of individual func-tions in establishing virtual organizations. Develop-ing the functional structure of the Intranet systeminvolves numerous variables and complexities thatare difficult to overcome objectively. This studybuilds a hierarchical structure of Intranet functionsfrom relevant literature and interviews with industryexperts. Based on the hierarchical structure of In-

tranet functions, this study conducted three surveysŽ .by using the Analytic Hierarchy Process AHP as a

systematic tool to evaluate the various functions ofthe Intranet. Three different groups of people—topmanagement, middle management, and low manage-ment—participated in the surveys because the rela-tive importance of Intranet functions may vary ac-cording to the organizational levels. It has beenargued that different types of information systemsshould be built for various organizational levels,because they require distinct types of informationw x11,12 . Therefore, top managers, middle managers,and low managers may not necessarily share thesame view regarding the relative importance of theIntranet functions. If they differ significantly, sys-tems developed for different target levels may haveto focus on varying functions.

The rest of this paper is organized into fivesections. Section 2 reviews the theory regardingvirtual organizations and the Intranet system thatforms the basis of the hierarchical structure of In-tranet functions. Section 3 explains the concept of

Ž .the Analytic Hierarchy Process AHP and presents ahierarchical structure of Intranet functions. Section 4explains the survey design and methods of dataanalysis, followed by Section 5 that presents theresults of the three surveys. Finally, this paper con-cludes with discussions on the implications of theresults.

2. Intranet for virtual organizations

2.1. Virtual organization

A virtual organization is defined as a group wherevarious independent companies link together to forma single temporary company, and where this linkage

w xundergoes transformations on a regular basis 4,8 .The individual companies in the virtual organizationbond with a common objective and share the samestrategies and processes.

A real world example of a virtual organizationcan be found in Verifone, a manufacturer of creditcard transaction equipment. The company has nohead office and its functions are spread around theworld. Verifone uses its IT infrastructure and com-munications systems to link together three manufac-

Page 3: Hierarchical Structure of Intranet Functions and Their Relative Importance:: Using the Analytic Hierarchy Process for Virtual Organizations

( )J. KimrDecision Support Systems 23 1998 59–74 61

w xturing centers located in different countries 30 . Allof Verifone’s internal communications are conductedvia electronic mail over the Internet. By using theInternet, Verifone is able to work around the clockon critical application developments. As one officegoes home, the code is passed onto another countryfor testing and documenting overnight. This bringssignificant time benefits to the company, and cus-

w xtomers are pleased with the responsiveness 30 .Other examples of well known corporations that aretaking advantage of virtual organizations are Hewlett

w xPackard 18 , IBM, AT&T, McDonald’s, Whirlpool,w xand Toyota 27 .

Virtual organizations have various advantagesover other organizational structures: it gives accessto a wide range of specialized resources; allowsindividual members to retain their independence andcontinue to develop their niche skills; eases thereshaping and changing of members according to theneeds of the project or task at hand; and eliminatesthe hassle of ‘divorce settlements’ as in formal joint

w xventures 29 . However, what makes a virtual organi-zation powerful may also make it vulnerable. Forexample, as a company connects with others to forma virtual organization, it faces the danger of exposingits secrets and undermining its core competence.Therefore, even while stressing the relations with itspartners, the individual company should also struggleto maintain its internal unity and core competencew x2 .

Such vulnerabilities had prevented the wide-spreadimplementation of virtual organizations prior to theemergence of the Intranet. The Intranet is believed tohave the potential to complement the weakness ofvirtual organizations for several reasons. First, work-ers located in different parts of the world can com-municate with one another in a timely and cost-effec-

w xtive manner 24 . Second, the Intranet system allowsusers to freely access external data as well as infor-mation within the organization, while securely pro-tecting their core competencies from outside intrud-

w xers 31 . Third, because of its comparatively inexpen-sive implementation and utilization costs, the In-tranet allows organizations to expand without incur-ring the same degree of relocation expenses, and alsolets them advertise and promote themselves to amuch wider audience at a lower cost than ever

w xbefore 24 .

2.2. Intranet

The Intranet is the implementation of Internettechnologies within an organization, to enhance theorganization’s efficiency by providing all availableresources to each employee’s desktop with minimal

w xcost and time 22 . It combines hardware and soft-ware to build a secure network that collects, pro-cesses, transmits, and categorizes information byutilizing a web browser as the user interface and

w xTCPrIP as the communication protocol 20 .Despite the still existing unknown’s and unavail-

able’s, Intranets are already producing high rewardsŽ .for corporations owing to several factors: 1 they

are cheaper, faster, and easier to set up than client-server network systems, given the use of existing

Ž .TCPrIP for outside communications; 2 their archi-Ž .tecture is already established and built into PC’s; 3

their platform offers room for growth and flexibility;Ž .4 they allow the company to have a very flexible

w x Ž .hardware and software strategy 14 ; and 5 theyrequire only a short time for the users to learn how

w xto use the system 10 . In short, because of theirexcellence in connectivity, Intranets make it mucheasier to communicate with other partners around the

w xworld at any time and at any place 30 .As a result of such benefits, Intranets are being

used to support a wide range of corporate functionssuch as document management, group scheduling,project management, and knowledge managementw x14 . In particular, Intranets are being heavily used inpublishing internal reports on human resources andcorporate policy, sharing corporate knowledge withteam members, and accessing data and information

w xfrom internal sources 19 . Results of a recent surveyindicate that users expect to expand their use of theIntranet and move their mission critical applicationsonto the Intranet, many of which will be active by

w x1998 3 .Despite the many functions of the Intranet, more

need to be developed in order for it to be used morew xwidely 20 . Current Intranet capabilities roughly

include electronic mail systems, online publishing,online searches, and application distribution pro-grams. In order to implement the Intranet in largecorporations, functions such as groupware-like tools,common user interfaces, wide area database access,and platform independent applications must be de-

Page 4: Hierarchical Structure of Intranet Functions and Their Relative Importance:: Using the Analytic Hierarchy Process for Virtual Organizations

( )J. KimrDecision Support Systems 23 1998 59–7462

veloped in addition to those currently available.Moreover, if organizations are to truly reap theadvantages of the Intranet, additional functions suchas file management systems, print services, and ap-

w xplication sharing programs need to be improved 20 .Since current Intranet systems already possess nu-merous functions and more are expected to be addedto future systems, it is essential to structure theoverall hierarchy of Intranet functions. Section 3provides a hierarchical structure of Intranet functionsby using the AHP methodology.

3. An analytic hierarchy of Intranet functions

3.1. OÕerÕiew of the AHP

Ž .The AHP Analytic Hierarchy Process can bedefined as a hierarchical analysis methodology sup-porting rational decision making by simplifying a

w xcomplicated problem 16 . In the AHP method, thefirst step is to set up the objective of decisionmaking. In this study, the objective is to develop anIntranet system that is ideal for a virtual organiza-tion. The primary goal is to determine what functionsshould be included in the ideal Intranet system andhow important the individual functions are for devel-oping virtual organizations. Therefore, the main ob-jective, ‘Ideal Intranet for a Virtual Organization’was placed at the top level of the analytic hierarchyshown in Fig. 1.

The second step is to break down the objectiveinto functional factors and organize them within an

Fig. 1. AHP: ideal intranet for a virtual corporation.

analytic hierarchy. In this study, the second level ofthe hierarchy was divided into two elements—Flexi-bility and Unity. These are the sub goals of theoverall goal of developing the ideal Intranet systemfor a virtual organizations. Similar decomposition ofthe functional elements was made iteratively down tothe fifth level, which contained the leaf functions of

Ž .the hierarchy e.g., platform . There is no singlecorrect hierarchy for a given system, and severaldifferent hierarchies can be built depending on dif-ferent perspectives. This study initially built an ana-lytic hierarchy based on relevant literature on virtual

w x w xorganizations 4 and the Intranet 20 . The initialhierarchy was revised based on the opinions of sixindustry experts—three Intranet system developersand three users—to ensure the comprehensivenessand comparability of the functions at each level ofthe hierarchy. The final hierarchy, shown in Fig. 1,will be explained in the subsequent sections.

Once the analytic hierarchy has been constructed,the next step is to compare the factors at the samelevel in pairs and measure the comparative contribu-tion of the factors to the main objective. This pair-wise comparison enables the decision maker to eval-uate the contribution to the objective of each factorindependently, thereby simplifying the decision mak-ing process. In this study, three groups of respon-dents performed the pairwise comparisons, and Ex-pert-Choice was used to convert the results of thepairwise comparison into the relative importance ofindividual functions. Details about the surveys andtheir results will be explained later in Section 4.

There are several reasons why the AHP is suitablefor selecting the ideal Intranet for a virtual organiza-tion. First, in order to establish an Intranet, it isessential to consider not only the technical factors,but also qualitative factors such as objectives ofvirtual organizations. The AHP method enables themanagers to make qualitative decisions more objec-tively via the quantitative comparison of qualitativefactors. Second, the AHP supports systematic deci-sion making by expressing the interaction and hierar-chy of factors, thus reducing the danger of a roughestimation. Third, since the AHP describes trade-offrelations between factors, it allows to consider thebalance of factors at the same level. Finally, AHPprovides a framework for group participation in deci-

w xsion making or problem solving 16 . The group

Page 5: Hierarchical Structure of Intranet Functions and Their Relative Importance:: Using the Analytic Hierarchy Process for Virtual Organizations

( )J. KimrDecision Support Systems 23 1998 59–74 63

participation is important because members of theŽvarious groups such as the top management, middle

.management, and low management of the organiza-tion should be present when making crucial deci-sions such as developing the Intranet system. Thefollowing sections explain in detail the individualfunctions that are expressed in Fig. 1.

3.2. Flexibility and unity

Since a virtual organization is a composition ofdifferent individuals, flexibility and unity are the twomost critical factors for its successful implementa-tion. The analytic hierarchy in Fig. 1 reflects theimportance of flexibility and unity by drawing themas the two entities directly below the top goal.

Flexibility means that a virtual organization shouldbe able to adapt itself readily to the changing envi-ronment. Virtual organizations exist only to serve aspecific need, and the team members are carefullychosen to meet this need and no other. The teammembers come from anywhere in the world based onspecific competencies such as technical expertise or

w xdomain specific knowledge 24 . Upon completion ofa certain project, the team disperses. Thus, the In-tranet should support the flexible gathering and dis-

w xpersion of members 4,22 .Unity signifies the integration of independent

companies that compose a virtual organization. Themembers of virtual organizations should respond au-tonomously and readily to changing customer de-

w xmands and the environment 29 ; this autonomy eas-ily leads each party in a virtual organization to have

w xits own goal and to act in its own self-interest 2 .However, to manage a virtual organization success-fully, all participants must still behave as if they are

w xone organization as in a traditional corporation 19 .In the absence of headquarters, it is quite difficult,yet important, to implant the sense of belonging inthe members’ minds. Therefore, one of the keyfunctions of the Intranet should be to foster the

w xseamless integration of operations 24 and the shar-ing of a common goal among the individual compa-

w xnies 23 .Therefore, Intranet functions that increase the

flexibility and unity of the organization are importantto achieve the ultimate goal of the Intranet system

for virtual organizations. However, it is very difficultto satisfy flexibility and unity at the same time

w xbecause they are in a trade-off relation 2 . As thevirtual organization becomes more flexible, it be-comes more difficult to maintain integration basi-cally because each company may run opportunisti-cally. While a virtual organization with maximumflexibility could quickly respond to the changingenvironment, the companies may have difficulties incoordinating activities with other members. On theother hand, a fully integrated virtual organization isslow in responding to the environment, but has lessdifficulty in coordinating members to go in the same

w xdirection 2 . Therefore, at the first level of theanalytic hierarchy, it is essential to establish a bal-ance between Intranet functions to foster flexibilityand unity.

3.3. Compatibility and adaptability for flexibility

Flexibility can be assessed by two factors: com-patibility and adaptability. Compatibility means theflexibility of the information infrastructure in a vir-tual organization that is continuously changing itself.One of the most significant expenses related to infor-mation systems is the cost of maintaining the com-patibility between existing old programs and newly

w xdeveloped ones 8 . This cost could produce seriousproblems especially in a virtual organization becauseof the organization’s need to continually change itsinformation infrastructure. Thus, the Intranet shoulddeliver a compatible information system to save timeand cost of maintenance.

Adaptability signifies the ability of a virtual orga-nization to modify its organizational structure ac-cording to the changing environment. Since it is anopportunity-pulled and opportunity-defined integra-tion of core competencies, the virtual organizationhas to be able to continuously adapt itself to the

w xchanging environment 7 . Therefore, the Intranetshould provide functions that enable the virtual orga-nization to respond agilely to the opportunities asthey arise.

Adaptability may have a trade-off relation withcompatibility. If a decision maker focuses too heav-ily on a compatible information system, it will behard to change or revise the information architecture,

Page 6: Hierarchical Structure of Intranet Functions and Their Relative Importance:: Using the Analytic Hierarchy Process for Virtual Organizations

( )J. KimrDecision Support Systems 23 1998 59–7464

and the alternatives of reengineering will be limited.On the other hand, if the organization concentratessolely on adaptability and changes its structure toofrequently, the information system will not be able tokeep up with the changing organization.

3.3.1. Compatibility of hardware and softwareThe compatibility of the information system is

essential for different independent companies to co-operate with each other on different hardware andsoftware. Table 1 shows the sub and leaf functions ofcompatibility and their evaluation criteria.

Hardware compatibility means that a worker whouses a Mackintosh computer, for example, should beable to communicate freely with others using an IntelPC. A hardware system consists of the platform andperipheral deÕices. The platform compatibility canbe assessed in terms of the number of platforms theIntranet system supports and the ease of a commoninterface among different platforms. The compatibil-ity of peripheral deÕices stands for the number ofdifferent types of such devices the Intranet can sup-

Ž .port e.g., fax machines, scanners, printers, etc. .The other compatibility required for a virtual

organization is software compatibility. Softwarecompatibility signifies that workers using differentsoftware should be able to share data and applica-tions freely with one another. Software compatibilityhas four leaf functions—the compatibility of applica-

tion programs, operating systems, data bases, andprogram development tools. Application compatibil-ity is evaluated by the number of programs that canbe executed together and by the degree of ease ininterrelating programs. OrS compatibility is mea-sured by the number of operating systems that areworking in the Intranet system and by the ease ofproviding a common interface among differentOrS’s. Data base compatibility is assessed by thenumber of data bases where data can be accessed,and by the ease of data access. Lastly, the compati-bility of program deÕelopment tools is assessed bythe number of API’s, by the ease of programming,and by the provision of debugging tools and applica-tion builders.

3.3.2. Adaptability of people and processAs in the compatibility of information systems,

adaptability can also be considered in two dimen-sions—people and process. Table 2 shows the suband leaf functions of adaptability and their respectiveevaluation criteria.

The term people in the AHP signifies the adapt-ability of the human members comprising a virtualorganization. Since a virtual organization is a unioncomprised of independent companies with variouscore competencies, it is important to find experts orcompetent people and deploy them to the right placeas needed. The people section can be divided into

Table 1Sub and leaf functions and evaluation criteria for compatibility

Compatibility

Sub functions Leaf functions Evaluation criteria

Hardware Platform Number of platforms supportedEase of providing a common interface among different platforms

Peripheral devices Number of different types of fax machines, scanners, printers supportedSoftware Application Diversity of compatible programs

Ease of exchanging programsOperation systems Number of operating systems available

Providing a common interface among different operating systemsData base Number of accessible data bases

Ease of data accessDevelopment tools Diversity of programming tools: number of APIs

Ease of programmingProvision of debugging tools and application builders

Page 7: Hierarchical Structure of Intranet Functions and Their Relative Importance:: Using the Analytic Hierarchy Process for Virtual Organizations

( )J. KimrDecision Support Systems 23 1998 59–74 65

Table 2Sub and leaf functions and evaluation criteria for adaptability

Adaptability

Sub functions Leaf functions Evaluation criteria

People External participants Ease of accessing expert advice and customer informationNumber of communication channels

Internal members Ease of providing members’ profiles reliably and quicklyProvision of intelligent task assignment system

Process Pre-arrangement Number of tools for modeling and simulating a processEase of manipulating a process before it starts

Runtime revision Ease of checking a process’s bottleneckProvision of process revision tools

two groups—external participants and internalmembers. External participants consists of outsidersinteracting with a virtual organization such as ex-perts and customers. Internal members are the em-ployees of an independent company. To utilize exter-nal people dynamically according to the changingenvironment, the Intranet should be able to providethe stakeholders with information and communica-tion channels. The leaf function of external partici-pants is evaluated by the ease of providing informa-tion, as well as the variety of communication chan-nels. Also, to utilize internal members effectively,the Intranet should be assessed in terms of theprovision of internal members’ profiles in a reliableway and the provision of intelligent systems whichassign and arrange tasks appropriately among mem-bers.

The other part of adaptability is process, whichmeans the ability of an organization to freely changeits work process. In order to adapt to the turbulentenvironment, a virtual organization should be able tochange its processes fluently, which is essential foradaptability, and furthermore, flexibility. Process hastwo leaf functions—pre-arrangement and runtimereÕision. Pre-arrangement means adjusting a pro-cess before it starts running, and runtime reÕisionmeans the capability of adjusting a process during itsexecution. Pre-arrangement is evaluated in terms ofthe number of tools for modeling and simulating aprocess, as well as the ease of manipulating a pro-cess before it starts running. Runtime reÕision isevaluated by the ease of checking bottlenecks in theprocess and the provision of tools that enable aworker to revise a process on the fly.

3.4. Internal and external unity

For the unity of a virtual organization, an inde-pendent company should maintain its internal andexternal unity. Internal unity is the unity within theindependent companies that comprises a virtual orga-nization. Most of the successful virtual organizationsthus far have nurtured and guarded their internal corecompetencies, and have invested a considerableamount of resources to maintain their internal unity

w xfor this purpose 2 . In the long run, a companywithout the internal unity to guard its unique corecompetency will lose its strategic position in a vir-

w xtual organization 2 . Therefore, the Intranet systemshould provide functions that support internal unityfor retaining one’s core competency.

External unity is the unity with the outer part ofan independent company. Since a virtual organiza-

w xtion is borderless 7 , the relationship with externalparticipants keeps changing and is hard to define.Therefore, it is necessary for a virtual organization tomaintain a firm connection with customers, govern-ment departments, and even competitors, as well asits internal partners. The Intranet system should pro-vide a function that supports to establish and main-tain firm connections with the external participants.

Internal unity and external unity may be in atrade-off relation. If an independent company con-centrates too much on its internal unity, the organiza-tional structure will be rigid. As a result, it will bedifficult for the company to coordinate activitieswith its partners and respond to the external environ-ment. On the other hand, if an independent companyinvolves itself too much with the outside world, it

Page 8: Hierarchical Structure of Intranet Functions and Their Relative Importance:: Using the Analytic Hierarchy Process for Virtual Organizations

( )J. KimrDecision Support Systems 23 1998 59–7466

increases the danger of being over-exposed to theexternal environment. Therefore, Intranet functionsshould support keeping a balance between its inter-nal and external unity.

3.4.1. Empowerment and control for internal unityFor internal unity, a company making up a virtual

organization should be able to empower and controlw xits employees 5,25 . Table 3 shows the sub and leaf

functions of internal unity and their evaluation crite-ria.

Empowerment is the function that provides work-ers with more abilities and authorities to make theirown decisions in order to respond to the environment

w xeffectively 15 . In order to keep its core competencein a time-based competition, a company should es-tablish an environment where workers could escalatecreativity and innovation. Intranet functions to sup-port empowerment enables workers to communicatewith each other and to access appropriate data asmuch as possible. The leaf function of communica-tion is assessed by the diversity of communicationchannels, such as video conferencing and bulletinboards, and the ease of using such channels. The leaffunction of data access is measured by the numberof data bases accessible, by the reliability of databases, and by the ease of handling non-formalizeddata.

Control is another function with which managersmaintain the internal unity within an organization. It

is important for managers to contain their workers,because not doing so could cause a conflict ofactivities. The manager of an independent companyshould be able to control the products, processes,and people as he or she had regulated them beforeforming part of a virtual organization. The functionof controlling product can be assessed by the easeof checking the cost, quality, and time of product, aswell as the ease of analyzing results and providingnecessary feedback. The function of controlling pro-cess and people can be assessed in a similar way asproduct.

3.4.2. ConnectiÕity and security for external unityIn order to support external unity, the Intranet

must provide two groups of functions—connectiÕityand security. Table 4 shows the sub and leaf func-tions of external unity and their evaluation criteria.

ConnectiÕity is the accessibility of external infor-mation. For connectiÕity with the outside of anindependent company, it is essential for the companyto escalate both the quantity and quality of informa-tion. Information quantity stands for how much in-formation internal members can freely access, andinformation quality signifies how effectively theycan find the right information needed. Informationquantity could be evaluated in terms of the numberof information channels, media, and interfaces. In-formation quality, on the other hand, could be as-sessed by the speed of information search and by the

Table 3Sub and leaf functions and evaluation criteria for internal unity

Internal unity

Sub functions Leaf functions Evaluation criteria

Empowerment Communication Provision of diverse communication channelsCompatibility of e-mail systemsCall of message arrival

Data access Number of data bases accessibleReliability of data baseEase of handling non-formalized data

Control Product Ease of checking cost, quality, and time of productsEase of result analysis and feedback

Process Ease of checking cost, quality, and time of processesEase of result analysis and feedback after process

People Ease of checking cost, quality, and time of peopleEase of result analysis and feedback

Page 9: Hierarchical Structure of Intranet Functions and Their Relative Importance:: Using the Analytic Hierarchy Process for Virtual Organizations

( )J. KimrDecision Support Systems 23 1998 59–74 67

Table 4Leaf functions and evaluation criteria for external unity

External unity

Sub-functions Leaf-functions Evaluation criteria

Connectivity Information quantity Number of information channelsDiversity of media and interfaces

Information quality Speed of information searchReliability of information category

Security External intrusion Provision of firewall, intrusion detection system, password and authenticationInternal misuse Provision of encryption and address translation

variety of tools categorizing data to get reliableinformation.

Security is the ability of an independent companyto protect its core competency and internal integra-tion. In a virtual organization environment, a com-pany can be subjected to serious risks of exposing itssecrets to the outside, so security becomes vital when

w xjoining a virtual organization 2 . Security can beconsidered in two ways. First, external intrusionmeans attacks by competitors or hackers, and thesecurity against external intrusions should be evalu-ated by the provision of firewall, an intrusion detec-tion system, passwords, and other authenticationtechnologies. Second, internal misuse means the out-flow of information by mistake, and Intranet func-tions against internal misuse can be evaluated by theprovision of encryption and address translation tech-nologies.

4. Survey design and data analysis

A survey was conducted to measure how peopleperceive the relative importance of the various func-tions proposed in the analytic hierarchy of the In-tranet explained in Section 3. As shown in Section 3,most Intranet functions are in trade-off relationswhere the level of support for one function cannot beincreased without decreasing the level of support foranother. In order to keep the balance between thefunctions with trade-off relations, we need to knowwhich are more important than others. This informa-tion guides us to focus the development efforts onthe important functions and spare extra efforts forthose that are perceived relatively less important.This section describes the survey design and the

procedure of data analysis that were conducted toidentify the relative importance of the Intranet func-tions.

4.1. Questionnaire design

Based on the analytic hierarchy in Fig. 1, aquestionnaire was developed in order to assess therelative importance of the Intranet functions. Theanalytic hierarchy consists of 15 sets of functions,where each set comprises of functions from the sameparent node at the same horizontal level. There are

Ž13 sets with two functions e.g., Hardware vs.. Žsoftware , one set with three functions product,

.process and people , and a set with four functionsŽapplication, ors, database, and programming

.tools . The sets with two functions needed only onepairwise question, while the sets with three and fourfunctions needed three questions each, resulting in atotal of 19 questions. Each question consisted of apairwise comparison of two elements at the samelevel of the hierarchy. For each question, the respon-dents were asked to indicate the relative importanceof two functions with respect to the parent node inthe hierarchy. For example, in the first question asshown in Fig. 2, the respondents were asked toindicate the relative importance of functions support-ing flexibility and unity when selecting the Intranetsystem that would maximize the effectiveness of avirtual organization.

The questions were ordered to follow the analyticŽhierarchy in a breadth-first pattern vertically from

top to bottom, and horizontally from left to right at.the same depth . Consequently, the first question

dealt with the relative importance between flexibilityand unity, and the last question concerned the rela-

Page 10: Hierarchical Structure of Intranet Functions and Their Relative Importance:: Using the Analytic Hierarchy Process for Virtual Organizations

( )J. KimrDecision Support Systems 23 1998 59–7468

Fig. 2. Sample question.

tive importance between external intrusion vs. inter-nal misuse. The breadth-first pattern was adopted inorder to focus the respondents’ attention on the moreabstract functions first before being distracted by thedetailed implementation issues. The phrasing of thequestions were formulated carefully in order to re-flect the proper relationship between the elements inone level with the property in the next higher level.

4.2. SurÕey participants

The participants in this survey were categorizedŽ .into three groups–Top Management TM , Middle

Ž . Ž .Management MM , and Lower Management LM .All three groups were comprised of students whowere enrolled in classes at Yonsei University.

The respondents of the TM group were enrolledin an executive MBA course. They consisted of 37participants whose profiles constitutes 27 membersas presidents of their own company and 10 as vicepresidents in large conglomerates. As executives oftheir respective organizations, most members of theTM group were in charge of hundreds of employees,some having as many as 60,000 subordinates. Somemembers were responsible for companies where theyearly total sales figure reached as high as USD$20billion. The TM group was included in this studybecause they are the final decision makers whodetermine whether or not the Intranet system shouldbe implemented in their organization.

The members of the MM group were takingcourses in a flex-time MBA program. The MMgroup was composed of 29 employees of a majorKorean conglomerate whose companies include vari-ous industries such as a passenger airline, a tiremanufacturing company, an importrexport com-pany, and a large-scale construction industry, among

others. All 27 members of the MM group had at least3 years of management experience, with some hav-ing as many as 20 years. The MM group wasincluded in this study because they carry out impor-tant decisions of Intranet projects within their respec-tive divisions. They were also considered importantbecause of their value as middle managers who actas mediators between the TM and LM groups.

The LM group was made up of 32 senior under-graduate students, majoring in business administra-tion, who plan on graduating a month after takingpart in the survey. Hence, they were assumed to be asuitable representation of the line workers in realbusiness organizations. All possessed basic knowl-edge on the Internet and computer networking, andthey would be the actual users of the Intranet system.

4.3. Data collection

The survey was conducted in three independentsessions within a span of 2 weeks. Most respondentswere familiar with the use of the Internet, because allof the surveys were given at the end of their respec-tive courses on information technology. In each ses-sion, the experimenter followed a script word forword with a minimum level of digression. The scriptwas used to ensure the consistency of the examiner’sexplanations and eliminate any biases that could becaused by different phrasing of the questions insessions that are weeks apart. The script was writtenby the authors and verified by a third party networkspecialist prior to being used in the surveys.

Each session lasted approximately 90 min. For thefirst 30 min the experimenter presented the basicprinciples of the Internet, Intranet, and virtual orga-nizations. After the presentation of the basic princi-ples, the experimenter explained the individual func-

Page 11: Hierarchical Structure of Intranet Functions and Their Relative Importance:: Using the Analytic Hierarchy Process for Virtual Organizations

( )J. KimrDecision Support Systems 23 1998 59–74 69

tions at each set of the analytic hierarchy. Theexperimenter only explained the meaning of eachfunction without providing any information about therelative importance of the individual functions. Afterexplaining about one set of functions, respondentswere asked to mark their opinions regarding therelative importance of Intranet functions on the ac-tual questionnaire. This was carried out one questionat a time, where the respondents were allotted 1 minto mark their opinions after the experimenter ex-plained a set of functions.

4.4. Data analysis

The analytic hierarchy of Intranet functions wasprogrammed into Expert-Choice—a decision supportsystem that implements the analytic hierarchy pro-

w xcess 6 . The marked responses from each of thesurvey participants were entered into the Expert-Choice program to calculate the relative importanceof the functions. Based on the relative importancegiven by individual participants, the mean values ofrelative importance for the Intranet functions werecomputed. For each set of functions, these meanvalues were compared between the three groups ofrespondents to identify any group differences in theirperceived importance. If no significant group differ-ence was found in a set of functions, the three groupswere considered as one large group, where the meanvalues for the relative importance were comparedbetween the functions in the set to identify anyunanimous preference. If there was a significantgroup difference, the three groups were checkedseparately to see whether there was a differencebetween the functions for each of the three groups.

The results of the data analysis are presented inSection 5.

5. Survey results

5.1. OÕerÕiew of surÕey results

According to the survey results, the fifteen sets offunctions were classified into three categories, asshown in Table 5. For seven out of fifteen sets offunctions, most respondents placed a significantlyhigher importance on one function compared to the

Ž .other function s , regardless of organizational level.For example, most respondents agreed that adapt-ability is significantly more important than compati-bility in developing Intranet systems. These setswere named ‘Unanimously Preferred’ functions inTable 5. For five other sets of functions, respondentsin different organizational levels varied in the degreeof importance they rated. An example can be foundin the set of flexibility vs. unity where TM and LMsignificantly preferred flexibility while MM did notshow any preference. These sets of functions werenamed ‘differently preferred’ in Table 5. Finally,respondents did not show any preference for theremaining three sets of functions, which were named‘Unanimously Indifferent’ functions in Table 5.

5.2. Unanimously preferred functions

Table 6 summarizes the seven sets of Intranetfunctions that were unanimously preferred regardlessof different organizational levels. The notation ‘X)

Y’ indicates that respondents in the survey consid-

Table 5Survey results overview

Category name Sets of groups

Unanimously preferred Compatibility vs. adaptability; hardware vs. software; application vs. ors vs. database vs. programming-tools;Ž .seven sets of functions pre-arrangement vs. run-time revision; empowerment vs. control; communication vs. data access; information

quantity vs. information qualityDifferently preferred Flexibility vs. unity; internal-unity vs. external-unity; connectivity vs. security; platform vs. peripheral-devices;Ž .five sets of functions external-participants vs. internal-membersUnanimously indifferent People vs. process; product vs. process vs. people; external-intrusion vs. internal-misuseŽ .three sets of functions

Page 12: Hierarchical Structure of Intranet Functions and Their Relative Importance:: Using the Analytic Hierarchy Process for Virtual Organizations

( )J. KimrDecision Support Systems 23 1998 59–7470

Table 6Unanimously preferred functions

Preference F value P value

Ž . Ž .Adaptability 57.8% )compatibility 42.2% 9.78 p-0.005Ž . Ž .Software 70.8% )hardware 29.2% 108.71 p-0.0001Ž . Ž . Ž . Ž .Database 34.2% )application 25.4% )OrS 21.3% )programming tools 19.1% 30.27 p-0.0001

Ž . Ž .Run time revision 59.8% )pre-arrangement 40.2% 17.14 p-0.0001Ž . Ž .Empowerment 61.9% )control 38.1% 25.95 p-0.0001Ž . Ž .Communication 57.3% )data access 42.7% 8.93 p-0.005

Ž . Ž .Information quality 60% ) information quantity 40% 16.35 p-0.0001

ered that Intranet functions supporting X were signi-ficantly more important than those supporting Y. Thefollowing discusses the seven sets of functions, oneby one, from top to bottom in Table 6.

In order to build an efficient virtual organization,the respondents in all three groups rated the abilityof organizations to readily adapt to the changing

Ž .environment 57.8% as more important than havingan information system that has a high compatibility

Ž . Ž .with other systems 42.2% Fs9.78, p-0.005 .This can be attributed to the recent surge in thediversity of customers’ demands caused by themass-customization trend. As consumers’ demandsbecome more diverse and frequent, a virtual organi-zation must be able to adapt its organizational struc-ture as soon as a new request is made. Hence,respondents may have perceived adaptability as be-ing more important. Meanwhile, an increasing num-ber of corporations are developing standard systemsthat are compatible with most other systems, espe-cially in the area of Internet applications. Therefore,even though compatibility by itself is an importantissue, it can be left up to the developers of computersystems, and the fact that not many systems expertsparticipated in this survey might explain the rela-tively low importance for compatibility.

Among the compatibility of hardware and soft-ware, increasing the compatibility of software pro-

Ž .grams 70.8% was regarded as having priority overŽ . Žincreasing hardware compatibility 29.2% Fs

.108.71, p-0.0001 . This may be attributed to thefact that ordinary users need to deal with the compat-ibility of software more frequently, while only lim-ited systems personnel are exposed to the problem ofhardware compatibility. Also, the respondents mayhave felt that software is more important than hard-

ware in general since hardware is useless withoutappropriate software. For software compatibility,

Ž .database 34.2% was seen as being the most impor-Ž .tant function, followed by applications 25.4% , op-

Ž .erating systems 21.3% , and programming toolsŽ . Ž .19.1% Fs30.27, p-0.0001 . This may be at-tributed to the fact that usually database is the finalproduct that people actually interact with, and OrSand applications are used not as often as database,while the programming tools are used only by sys-tems developers.

Among the functions to increase the adaptabilityŽ .of organization process, run time reÕision 59.8%

was regarded as being more important than pre-Ž . Ž .arrangement 40.2% Fs17.14, p-0.0001 . This

may be because respondents recognized the inherentdynamics involved in business processes of virtual

w xorganizations 17 . In the era of mass-customization,it is practically impossible to plan the optimal pro-cess for every case before the process is initiated. So,a more practical way may be to apply a templateprocess initially and fine tune the process dynami-cally as situations change.

In terms of unity, the Intranet must be able tounite various different companies as one in order tobring out its full potential. To accomplish this, anIntranet function that supports the empowerment of

Ž .the individual workers 61.9% was selected as beinga more important means of internal unity than afunction for increasing a manager’s controllabilityŽ . Ž .38.1% Fs25.95, p-0.0001 . This may be partlybecause the current movements of business processreengineering have placed a high importance onempowerment than control, and partly because In-tranet systems were basically more prone to theconcept of empowerment because of the intrinsic

Page 13: Hierarchical Structure of Intranet Functions and Their Relative Importance:: Using the Analytic Hierarchy Process for Virtual Organizations

( )J. KimrDecision Support Systems 23 1998 59–74 71

nature of Internet technology. This result is alsoconsistent with the fact that run-time reÕision wasconsidered more important than pre-arrangement,because empowerment should award the ability todynamically adjust ongoing processes onto the peo-ple who actually perform the processes.

Among the ways to increase support for empow-Ž .erment, enhanced communication 57.3% between

employees within an organization was seen as beinga more effective means than increased access to dataŽ . Ž .42.7% Fs8.93, p-0.005 . This may be becauseof the fact that seeking help from experts is a moreeffective way to obtain the right information thanconsulting databases. Moreover, virtual organizationsusually assume the team-based format in which effi-cient communication between members is importantfor the performance of the entire team.

Ž .Finally, the quality of information 60.0% wasseen as being more important for a virtual organiza-

Ž .tion’s connectivity than information quantity 40.0%Ž .Fs16.35, p-0.0001 . This is probably to controlinformation overflow that has resulted from the wideuse of the Internet. The Internet has made it rela-tively easier to get more information, while it hasbecome relatively more difficult to obtain the rightinformation. Thus, the Intranet functions to increasethe quality of information, not the amount, was givenpriority among over quantity among the survey re-spondents.

5.3. Differently preferred functions

Table 7 summarizes the five sets of Intranet func-tions for which different organizational levels ex-pressed different degrees of relative importance. Thenotation ‘X)Y’ indicates that respondents in thespecific organizational level expressed that Intranetfunctions supporting X were significantly more im-

portant than those supporting Y. The notation ‘XsY’ stands for the fact that no statistical difference ofthe relative importance was observed among thefunctions in the set.

Three respondent groups showed significantly dif-ferent patterns in their appraisal of the relative im-

Žportance between flexibility and unity Fs4.29,.p-0.01 . Further evaluation of the relative impor-

tance for each group revealed that only the TMŽ . ŽFs5.32, p-0.01 and LM Fs90.54, p-

. Ž0.0001 groups regarded flexibility TM: 59.3%,. ŽMM: 55.2% more important than unity TM: 40.7%,.MM: 44.8% , whereas the MM group was indifferent

between the two functions. The TM group may haveplaced more importance on flexibility over unitybecause they want to respond to new business oppor-tunities as soon as possible. The LM group may haveshown a similar preference, because they like auton-omy in their work which is usually offered to themthrough flexibility. On the other hand, the MM groupput more importance on unity compared to the othertwo groups, because they have to control everydayoperations to which unity is as important as flexibil-ity.

For a virtual organization to unite and act as oneentity, a balance between unity among the differentcompanies within the organization and unity with theexterior of that corporation must be met. Results ofthe survey data revealed that there were significantdifferences among the three groups when rating therelative importance between internal unity and ex-

Ž .ternal unity Fs4.81, p-0.01 . Only the LMŽ .group preferred external unity LM: 74.0% over

Ž .internal unity LM: 26.0% in a statistically signifi-Ž .cant way Fs2.14, p-0.0001 . The other two

Ž .groups regarded external TM: 50.2%, MM: 56.6%Ž .and internal TM: 49.8%, MM: 43.4% unity as

being equally important. This can be attributed to the

Table 7Differently preferred functions

Sets of functions TM MM LM

Ž . Ž .Flexibility F vs. unity U F)U FsU F)UŽ . Ž .Internal unity IU vs. external unity EU IUsEU IUsEU IU-EUŽ . Ž .Connectivity C vs. security S CsS C)S C)S

Ž . Ž .Platform PL vs. peripheral devices PD PD)PL PDsPL PD-PLŽ . Ž .Internal members IM vs. external participants EP IM)EP IMsEP IM-EP

Page 14: Hierarchical Structure of Intranet Functions and Their Relative Importance:: Using the Analytic Hierarchy Process for Virtual Organizations

( )J. KimrDecision Support Systems 23 1998 59–7472

fact that the LM group members felt less of anattachment towards their respective organizations be-cause they have never been in charge of any organi-zation unit.

In enhancing the external unity of a virtual orga-nization, an Intranet system must keep a fine balancebetween maintaining connections with the outsideworld and guarding company secrets against externalthreats. The three groups showed significantly differ-ent patterns in their appraisal of the relative impor-

Žtance between connectiÕity and security Fs5.60,.p-0.005 . The TM group showed no statistically

Ž .significant preference Fs1.90, ns , while the MMŽ . Žgroup Fs8.60, p-0.005 and LM group Fs

.153.77, p-0.0001 displayed strong preferences forŽ .connectivity MM: 62.7%, LM: 80.2% over security

Ž .MM: 37.3%, LM: 19.8% . The MM and LM groupsmay have put more emphasis on connectiÕity be-cause they would not have enough external connec-tivity without the Intranet, while the TM group hasalternative ways to reach the outside world.

As for the three sets of functions just explained,although some groups expressed a significant prefer-ence while other groups did not, all three groupsagreed on the direction of their preferences. On theother hand, for the last two sets of functions in Table7, the three groups differed in terms of the directionof preference. This implies that some groups preferX over Y while other groups prefer Y over X both ina statistically significant way. The two sets are plat-form vs. peripheral deÕices and internal membersvs. external participants as shown in Table 7.

In order to support the compatibility of hardwarewithin a virtual organization, an Intranet system mustinclude functions to support the compatibility ofplatform and peripheral deÕices. The three respon-dent groups conveyed significant differences in com-paring the importance between the two functionsŽ .Fs7.19, p-0.001 . The TM group recognizedthe importance of supporting peripheral deÕicesŽ . Ž . Ž64.0% over platforms 36.0% Fs15.17, p-

.0.0001 , the MM group showed no preference be-Ž .tween the two Fs0.44, ns , while the LM group

Ž .rated the support of platforms 63.3% over periph-Ž . Ž .eral deÕices 36.7% Fs14.06, p-0.0005 . The

TM group may have put more emphasis on theperipheral deÕices because they are the mediumthrough which information is actually presented to

them. Meanwhile, the LM group may have regardedplatforms more important because performance ofthe group is more sensitive depending on the plat-forms rather than on the peripheral devices.

Finally, a key in adapting to the constantly chang-ing environment is the people who are involved invirtual organizations. The Intranet system shouldprovide both a function to support adaptation ofmembers within a certain virtual organization and afunction to support adaptation of participants outsidethe organization. Survey results showed that therewere statistically significant discrepancies among thethree groups in rating the relative importance ofinternal members vs. external participants. The TMgroup cited the importance of internal membersŽ .70.4% as being more essential than external partic-

Ž . Ž .ipants 29.6% Fs33.56, p-0.0001 , the MMŽgroup conveyed no statistical significance Fs3.25,

.ns , while the LM group preferred participants out-Ž .side an organization 73.4% over internal members

Ž . Ž .26.6% Fs46.52, p-0.0001 . This may be at-tributed to the fact that members at different organi-zational levels can differ in their definitions of theboundary of an organization. The TM group mayconsider the entire company as their organization,giving them a large pool to recruit personnel withdiverse expertise. On the other hand, the LM groupmay regard the department they were assigned to astheir organization, so the size of their organizationcan be very small. This may have led them to feelthat they need more external participants to be in-cluded in the virtual organization.

5.4. Unanimously indifferent functions

For three out of the 15 sets of functions, surveyresults showed neither any group difference amongthe three organizational levels, nor any relative im-portance among the functions within the sets. Thethree sets of functions are-adaptability of people vs.process, controllability of people vs. process vs.product, and prevention of external intrusion vs.internal misuse. The results imply that those func-tions in the three sets are equally important forvirtual organizations, and therefore, no group pre-

Ž .ferred one function over the other function s in thesets.

Page 15: Hierarchical Structure of Intranet Functions and Their Relative Importance:: Using the Analytic Hierarchy Process for Virtual Organizations

( )J. KimrDecision Support Systems 23 1998 59–74 73

6. Conclusion

The purposes of this study were to construct ananalytic hierarchy of Intranet functions and to iden-tify the relative importance of the functions for avirtual organization. The analytic hierarchy was ini-tially built based on the relevant literature in virtualorganizations and Intranets, and verified by industryexperts. The final hierarchy consisted of nineteensets of functions with a depth of four levels startingwith flexibility vs. unity at the top. Most of thefunctions in the sets are in trade-off relations wherefinding out the relative importance between func-tions is essential for the successful implementationof the Intranet. Three sessions of surveys were con-ducted with three groups of respondents at differentorganizational levels in order to identify the differ-ence in the relative importance of the Intranet func-tions. The results of this study indicate that the threerespondent groups agreed on the relative importanceof Intranet functions for seven out of fifteen sets. Onthe other hand, for the other five sets of functions,the three groups expressed different degrees andrordirections of preferences. Finally, the remaining threesets of functions were considered equally importantwith no group difference.

The implications of this study for the unani-mously preferred functions are evident. Developmentefforts should be focused on the functions that areconsidered relatively more important, rather thanbeing focused on the less important functions in thesame set. For example, respondents unanimouslyagreed that the function supporting the adaptabilityof organizations is relatively more important than thefunction supporting the compatibility of informationsystems. Therefore, when choosing between two In-tranet systems that are similar in all aspects, thesystem that has a higher support for organizationaladaptability should be selected over the one with ahigher support for compatibility. For another exam-ple, the survey results indicate that respondents eval-uated the function of dynamically adjusting pro-cesses as being more important than the function ofpre-arranging processes. Several modeling toolshave been developed to support run-time revisionsw x17 , including the effort to develop a modelingformalism for supporting the dynamic adjustment of

w xbusiness processes 9 .

Compared to the unanimously preferred functions,we need to be more cautious on the implications ofthe differently preferred ones. Special attentionshould be paid on the two sets of Intranet functionstowards which respondents at different organiza-tional levels expressed different directions of prefer-ences. This implies that functions of an Intranetsystem should be customized in order to conform tothe preferences of target users. For example, as ameans of adapting to the changing environment, theTM group placed a greater emphasis on the adapt-ability of internal members over the adaptability ofexternal participants, the MM group showed nopreference between the two functions, while the LMgroup rated the function supporting external partici-pants as being more essential. Consequently, if anIntranet system is to be implemented in a mid-sizedcompany where there are only a few low-level man-agers, the functions that support the adaptability ofexternal participants need to be strengthened. On theother hand, for large conglomerates where numerousmanagers are in charge of a vast number of employ-ees, an Intranet system that supports the adaptabilityof internal members needs to be considered.

One of the main limitations of this study comesfrom the composition of the survey respondents. Thisstudy assumes that the senior undergraduate studentswould have had the same preferences as line work-ers, an assumption that has not been verified yet.Also, all three groups did not contain a large numberof information systems experts who may have haddifferent preferences from ordinary users. An addi-tional survey is planned in order to identify therelative importance which actual line workers andinformation systems experts perceive towards thevarious functions of Intranet systems.

Acknowledgements

The author would like to thank Korean Depart-ment of Information and Communications for pro-viding a generous grant. He also thanks the twoanonymous reviewers of Decision Support Systemsand members of the Cognitive Information Engineer-ing Laboratory at Yonsei University for their com-ments on the earlier draft of this paper.

Page 16: Hierarchical Structure of Intranet Functions and Their Relative Importance:: Using the Analytic Hierarchy Process for Virtual Organizations

( )J. KimrDecision Support Systems 23 1998 59–7474

References

w x1 R. Bernard, The Corporate Intranet, Wiley, New York, 1996.w x2 H.W. Chesbrough, D.J. Teece, When is virtual virtuous:

organizing for innovation, Harvard Business Review,Ž .JanrFeb 1996 65–68.

w x3 D. Coleman, Collaborating on the Internet and Intranets,Proceedings of the Thirtieth Annual Hawaii InternationalConference on System Science, Vol. 2, 1997, pp. 350–358.

w x4 W.H. Davidow, M.S. Malone, The Virtual Corporation,Harper Business, New York, 1992.

w x5 J.D. Duck, Managing change: the art of balancing, HarvardŽ . Ž .Business Review 71 6 1993 109–118.

w x6 Expert Choice, Expert Choice—Decision Support Software,Version 9.0., 1995.

w x7 S.L. Goldman, R.N. Nagel, K. Preiss, Agile Competitors andVirtual Organizations, International Thomson Publishing,New York, 1994.

w x8 R. Grenier, G. Metes, Going Virtual: Moving Your Organiza-tion into the 21st Century, Prentice-Hall, NJ, 1995.

w x9 H. Hahn, J. Hahn, J. Kim, A Cognitive Engineering Study onthe Development of an Object-Oriented Process ModelingFormalism, Proceedings of the Thirtieth Annual Hawaii In-ternational Conference on System Science, Vol. 4, 1997, pp.199–209.

w x10 B. Hedin, F. Fallstrom, V. Ionesco, An Intranet Solution fora Real time GPMS in Newspaper Production, Proceedings ofthe Thirtieth Annual Hawaii International Conference onSystem Science, Vol. 4, 1997, pp. 320–328.

w x11 K. Laudon, J. Laudon, Management Information Systems,Macmillan, New York, 1996.

w x12 W. Orlikowski, J. Baroudi, Studying information technologyin organizations: research approaches and assumptions, Infor-

Ž . Ž .mation Systems Research 2 1 1991 122–145.w x13 B.J. Pine, II, Mass Customization, Harvard Business School

Press, Boston, 1993.w x14 J. Quelch, L. Klein, The Internet and International Market-

ing, Sloan Management Review, 1996, Spring, pp. 60–75.w x15 M. Ramage, Engineering a smooth flow?—A Study of work-

flow software and its connections with business processreengineering, Working paper, University of Sussex, London,1994.

w x16 T.L. Saaty, Decision Making for Leaders—The AnalyticHierarchy Process for Decisions in a Complex World, RWSPublications, Pittsburgh, 1995.

w x17 R. Spek, H. Sol, Multimodelling, A Group Approach toModeling Interaction Within Intra-organizational Systems,Proceedings of the Thirtieth Annual Hawaii InternationalConference on System Science, 1997, Vol. 2, pp. 179–188.

w x18 CALS, Editorial—The War of Transition, 1996, availablefrom http:rrcals.debbs.ndhq.dnd.carenglishrbulletinris-sue03reditorial.html.

w x19 CBDC, Are Your Company’s Assets Virtual or Real? May31, 1996, available from http:rrcivic. netrcambridgecivic-networkrdougscolumn16.html.

w x20 R. Chellappa, A. Barua, A. Whinston, INTRANETS: Look-Ž .ing Beyond Internal Corporate Web Servers! 1996 , avail-

able from http:rrecworld.utexas.edurejourintrint.html.w x21 Forbairt, The Virtual Corporation and The Global Village

Effect, Forbairt Internet Report, available fromhttp:rrwww.nua.ierinternetrANewVisionrVirtualVillage.html.

w x22 G V T , T he In tranet, 1996 , ava ilab le fromhttp:rrmidir.ucd.ier ;bmacgettrIntranet.html.

w x23 C. Handy, Trust and the Virtual Organization, Harvard Busi-ness Review, MayrJune 1995, p. 40, available fromhttp:rr138.182.252.121rRPIrefr214a.htm.

w x24 J. Maloff, The Virtual Corporation, 1995, available fromhttp:rrpubs.iworld.comriwonlinerJul95rfeat46.htm.

w x25 T.W. Malone, Inventing The Organizations Of The 21stCentury: Control, Empowerment, And Information Technol-o g y , 1 9 9 6 , a v a ila b le f ro m h ttp :r r w w w .hbs.harvard.edurmisrmultimediarlinkrsmalone.html.

w x26 B. McKern, International Network Corporations in a GlobalEconomy, 1994, available from http:rrwww.gsia.cmu.edurafsrandrewrgsiarboschrmckernr94-9.htmatab1.

w x27 R.L. Nolan, H. Galal, C. Tuller, Virtual Offices: New Ar-rangements for Redefining Organization Boundaries, 1996,available from http:rrwww.hbs.harverd.edurmisrmulti-mediarlinkrsnolan.html.

w x28 Roberts, Intranet: Just a Trendy New Name for a LAN? WebWeek, Volume 2, Issue 8, June 17, 1996, available fromhttp:rrwww.webweek.comr96Jun17rIntranetrtrendy-LAN.html.

w x29 T.L. Skyrme, The Virtual Corporation, 1996, available fromhttp:rrwww.hiway.co.ukrskyrmerinsightsr2virtorg.htm.

w x30 R. Vadon, PCW CUTTING EDGE-Virtual Companies, VNUB usiness P ub lications, 1996 , availab le fromhttp:rrwww.vnu.co.ukrhcrpcwrcut2;2.htm.

w x31 D. Vine, Bending Space and Time: The Virtual Organization,Internet World, May, 1995, available from http:rrwww.research.umbc.edur ; warrrvirtog.html.

Jinwoo Kim is Assistant Professor in the Department of BusinessAdministration and a member of Cognitive Science ResearchCenter at Yonsei University, Korea. His primary research interestsfocus on the human-computer interaction issues in Cyber-space.His current areas of research are: visual reasoning with multiplediagrams in inter-enterprise engineering, emotive interface engi-neering in electronic commerce, shared knowledge base in thedistributed dynamic decision making with intranet, and cogni-tively efficient market structure for cyber shopping malls.