toward improved community- supporting systems design

Copyright © 2005, Idea Group Inc. Copying or distributing in print or electronic forms without writtenpermission of Idea Group Inc. is prohibited.

International Journal of Technology and Human Interaction, 1(4), 19-36, October-December 2005 19

Toward Improved Community-Supporting Systems Design:

A Study of Professional Community Activity

Malte Geib, University of St. Gallen, Switzerland

Christian Braun, University of St. Gallen, Switzerland

Lutz Kolbe, University of St. Gallen, Switzerland

Walter Brenner, University of St. Gallen, Switzerland

ABSTRACT

In this article, we analyze the design factors of community systems in two real-world professionalcommunities — a learning network and an expert network — that employ a mix of communicationmodes, that is, face-to-face communication and computer-mediated communication. Our objectivesare to determine which design factors influence community activity and therefore communityoutput. We furthermore intend to make recommendations to improve the design of communitysystems that support professional communities using a mix of communication modes. Our study isexploratory and based on action research given the lack of studies on the design of community-supporting systems in professional communities that employ a mix of communication modes. Toillustrate similarities and to enhance the generalizability of our findings, we analyzed two real-world professional communities in-depth, namely, a learning network and an interorganizationalexpert network. Our study shows that face-to-face communication is the primary mode ofcommunication in these communities; the community systems that they employ only have asupporting function. This leads us to a few design guidelines for the systems that support suchcommunities. Generally, community systems have to support professional communities’ workprocesses and relationship development. Important functions for work-process support are thosethat support face-to-face meetings (for the preparation and wrap-up of meetings) and that explicitlysupport specific work processes. Important functions for relationship development are functionsthat enable or facilitate face-to-face meetings, for example, member profiles.

Keywords: community-supporting system; community system; expert network; learningnetwork; professional community; system design; virtual community

INTRODUCTION

In recent years, it has become nor-mal to support geographically dispersedcommunities with advanced forms of com-

puter-mediated communication (CMC) sys-tems, usually based on Internet technology.These community-supporting systems (inshort, community systems), frequentlytermed teamware (Schulte, 1999) or

This paper appears in the journal International Journal of Technology and Human Interaction edited by BerndCarsten Stahl. Copyright © 2005, Idea Group Inc. Copying or distributing in print or electronic forms without writtenpermission

701 E. Chocolate Avenue, Suite 200, Hershey PA 17033-1240, USATel: 717/533-8845; Fax 717/533-8661; URL-http://www.idea-group.com

��

IDEA GROUP PUBLISHING

20 International Journal of Technology and Human Interaction, 1(4), 19-36, October-December 2005


groupware (Bach, Vogler, & Österle, 1999),support the interactive exchange and cre-ation of documents, online discussions, chatrooms, and role-based personalization. Be-sides communities that rely solely on CMC(frequently termed virtual communities;Rheingold, 1998), the majority of professionalcommunities employ a mix of CMC andother communication modes (i.e., telephone,fax, face to face). Much research has beendevoted to the analysis of virtual communi-ties (e.g., Bieber et al., 2002a; Bieber et al.,2002b; Godio, 2000; Rheingold) and to thecomparison of CMC with other communi-cation modes (e.g., Etzioni & Etzioni, 1999;Wiesenfeld, Raghuram, & Garud, 1999).However, there has been little research onthe design of community-supporting systemsin professional communities that employ amix of communication modes.

To address this gap, the objective ofour research was to analyze communitysystem design factors in professional com-munities such as learning and expert net-works that employ a mix of communica-tion modes. We addressed the followingresearch questions in detail:

1. Which community system design fac-tors influence community activity andtherefore community output?

2. How should community systems support-ing professional communities be de-signed?

Because our research was explor-atory, we used an action research (AR)approach (Checkland & Holwell, 1998).Action research is often used in the infor-mation-systems domain for the exploratoryanalysis of systems design in real-world set-tings (Davison, Martinsons, & Kock, 2004;Mansell, 1991).

Two real-world communities were theobject of our in-depth study: a learning net-work of postgraduate students and aninterorganizational expert network consist-ing of experts from different companiesworking in the areas of customer-relation-ship management (CRM) and knowledgemanagement (KM).

In the following section, we describethe research model developed from the lit-erature on computer-mediated communi-cation and virtual communities that whichpresents the causal relationships discoveredin previous research relevant to our re-search questions. Next, we describe ourresearch methodology. Subsequently, wedescribe and discuss the results of our re-search to arrive at propositions for the de-sign of community systems supporting pro-fessional communities. Finally, we summa-rize our findings and discuss further re-search opportunities.

THEORETICALBACKGROUND

Professional CommunitiesThe notion of community is a

socioscientific collective term for a specifictype of social group (Poplin, 1979; Sutton &Munson, 1976). Although there is no gener-ally accepted definition, a community canbe defined as a group of socially interactingpersons who are mutually tied to one an-other and regularly meet at a common place(Hillery, 1955). With the diffusion of elec-tronic information and communication sys-tems, communities have increasingly turnedto computer-mediated communication.

In respect to their objectives andscope, communities using CMC can beclassified into three major types (Markus,2002). Socially oriented communities formto support the development of social rela-



tionships between individuals and have noeconomic goal. Commercially orientedcommunities form to directly support aprofit-oriented economic goal (cp Hagel &Armstrong, 1997). Professionally orientedcommunities (in short, professional com-munities) consist of companies’ employeeswho communicate and share informationto support their professional tasks (Godio,2000). In this article, we focus on the analy-sis of professional communities.

Professional communities can be dif-ferentiated into expert networks and learn-ing networks (Markus, 2002). Expert net-works are formed by experts focused on aspecific topic with the aim of acquiring anddeveloping knowledge through their mutualinteractions and discussions as membersof the network. An expert network emergesthrough voluntary association and may beeither intra- or interorganizational. Asocioscientific explanation of the expert-network phenomenon is given by Wenger(1997), who calls this community type a“community of practice.” Wenger,McDermott, and Snyder (2002) offer apragmatic definition of expert networks as“groups of people who share a concern, aset of problems, or a passion about a topic,and who deepen their knowledge and ex-pertise in this area by interacting on an on-

going basis.” Learning networks areformed by professionals with the objectiveof joint education, training, or learning(Markus, 2002). Examples of these are vir-tual corporate universities (Kraemer &Müller, 1999).

Community SystemsWith the propagation of the Internet,

most CMC systems became based onInternet technology. We use the term com-munity system to describe the Internet-based application system that enables com-munity members to interact with one an-other.

To structure our analysis of commu-nity systems for professional communities,we use a research model that describes therelationships between community systemdesign factors and community activity preva-lent in the literature on CMC (see Figure 1).

In the following sections, we definethe concepts used in the research modeland explain the relationships between them.

Process SupportCommunities use community systems

to support their communication-based pro-cesses (Bieber et al., 2002a; Watson-Manheim & Belanger, 2002). These can beroughly divided into relationship-develop-

Figure 1. Research model

Activity

Satisfaction

Process Support

Work Processes

Relationship Development

Usability

Technology/Media Choice Training

Experience

Org. Culture



ment processes and work processes (ortask-oriented processes). In relationship-development processes, community mem-bers establish relational intimacy by ex-changing social information (Chidambaram,1996). In contrast, work processes are ex-ecuted by community members to solve aproblem or to work on a specific task. Ex-amples of work processes are coordina-tion, information gathering, knowledge shar-ing, conflict resolution, negotiation, and in-formation dissemination (Watson-Manheim& Belanger).

Technology and Media ChoiceIn the design of community systems,

the choice of technology and media dependson the work processes and relationship-development processes that they shouldsupport (Stanoevska-Slabeva & Schmid,2000; Watson-Manheim & Belanger,2002). Well-established technology is alsooften chosen on the basis of previous ex-perience with this technology. The choiceof technology may therefore depend on thecommunity members’ previous technologyexperience if one assumes that they influ-ence the design process (Chidambaram,1996; Watson-Manheim & Belanger). Thechoice of technology may also depend onthe organizational culture (Watson-Manheim & Belanger; Wiesenfeld et al.,1999). Subject to their previous experienceand choice of technology, community mem-bers require training in using the commu-nity system to ensure that it is effectivelyused (Wiesenfeld et al.). A communitysystem’s usability determines whetherpeople will find the system easy to use(Bieber et al., 2002a). This is primarily in-fluenced by the choice of a suitable tech-nology and users’ experience with the cho-sen technology.

Satisfaction, Activity,and Performance

Community members’ satisfactionwith the community system is influencedby the degree of process support offeredby the system (Chidambaram, 1996). Thesystem’s usability plays an important rolein the degree of satisfaction experienced.The community members’ activity is de-fined as the frequency with which com-munity members use the system for com-munication with one another. Activity isfurthermore primarily dependent on thecommunity members’ satisfaction with thesystem and also influenced by organiza-tional culture. For example, Hiltz andJohnson’s (1990) study showed that the bestpredictor of satisfaction with CMC was the(virtual) activity among group members.Experience’s (indirect) effect on activity isacknowledged by Chidambaram, who saysthat “experience with the medium can af-fect the extent of use…of the medium.”

According to McDermott (2002), ac-tivity — included in our research model (Fig-ure 1) — is the most basic concept by whichwe can measure communities’ perfor-mance. Performance can generally be mea-sured on different levels, each with a differ-ent impact on business result (see Figure 2).

Activities comprise, for example, meet-ings, discussions, and one-to-one contacts.Measuring activities can be helpful in givingsome indication of the communities’ health.However, these measures do not demon-strate the communities’ contribution to itsmembers or organizations. To determine thiscontribution, it is necessary to measure com-munities’ performance in terms of output andvalue (McDermott, 2002).

In focusing on the community sys-tems’ contribution to the overall perfor-mance of communities, we restrict perfor-mance measurement to the activity level.This is necessary because the community



systems’ goal is to facilitate the commu-nity members’ activities and to make themindependent of restrictions imposed byspace and time (Bieber et al., 2002a). Com-munity systems’ impact is therefore re-stricted to the activity level as the systemsonly influence business results by facilitat-ing activities.

METHODOLOGYGiven the lack of studies on the design

of community-supporting systems in profes-sional communities that employ a mix ofcommunication modes, our study is explor-atory and the research sites are real-worldcommunities. To illustrate the similarities andto enhance the generalizability of our find-ings, we made an in-depth analysis of twoprofessional communities, namely, a learn-ing network and an interorganizational ex-pert network.

Action research (AR) is an appropri-ate methodology for the exploratory analy-sis of systems design in real-world settings(Davison et al., 2004; Mansell, 1991). ARfocuses on solving organizational problemsthrough intervention, while at the same time

contributing to scholarly knowledge(Davison et al.). In the AR process, theresearcher enters a real-world situation andbecomes involved as both participant andresearcher (Checkland & Holwell, 1998).AR’s iterative characteristic implies a cy-clic process of intervention, with one orseveral cycles of activities being conducted(Davison et al.).

Checkland and Holwell (1998) arguethat in order for AR results to be valid, theresearch process has to be recoverable byinterested outsiders. It is therefore essen-tial to state the epistemology (the set ofideas and the process in which these ideasare used methodologically) through whichoutsiders make sense of the research, andthus define what they regard as acquiredknowledge (Checkland & Holwell). Figure3 shows the AR process that we followedfor one cycle in respect to each of the twocommunities.

The organizational problem to besolved in respect to each professional com-munity was improving the community’swork and relationship-development pro-cesses through the design of a community-supporting system.

Figure 2. Performance measurement framework for communities (McDermott, 2002)



A detailed understanding of the sur-rounding environment is a prerequisite forthe determination of an appropriate inter-vention, therefore the data-collection tech-niques employed before, during, and afterthe action-taking stages should ensure a richpool of data for subsequent analysis(Davison et al., 2004). Furthermore, a thor-ough review of the existing literature is use-ful not only to inform the research’s focusand process, but also to help position theresearch within scholarly knowledge(Davison et al.). We therefore developeda research model from the literature (seeFigure 1).

In the diagnosis phase, we started ourresearch with a thorough diagnosis of thecommunity’s current situation. This wasdone through an analysis of the community’sdocumentation and interviews with commu-nity members regarding their requirements.Subsequently, we identified the communitymembers, their roles, and their work andrelationship-development processes. We alsogained insights into the choice of appropri-ate technologies, media, and functions sup-porting these processes. Finally, we alsoacquired information on community mem-bers’ experience and training, as well as ontheir general cultural context.

In AR, the diagnosis will directly in-form the planning of actions, and planned

actions will subsequently be implementedand evaluated. In the action-planning phase,we planned the design of the communitysystem according to the users’ requirementsas based on their processes. In the inter-vention phase, we developed and imple-mented the community system.

In the evaluation phase, we comparedthe intervention outcomes with the projectobjectives and expectations. We thereforegathered performance data that were rel-evant within our research model’s contextto measure the community system’s suc-cess. This was done by measuring the com-munity members’ activities when using thecommunity system. The starting point ofthis data collection was the access logsgenerated by the Web server (LotusDomino server) that handles the commu-nication between the community membersand community applications (Lotus Notesdatabases, Lotus Team Workplace, andLotus Sametime). The analysis steps wereas follows:

1. We collected the access logs, which re-corded page views. A page view is theresult of a request for a particular Webpage and therefore denotes the request-ing person’s activity.

2. To eliminate the effects produced by acommunity system’s administration

Figure 3. AR process model (Checkland & Holwell, 1998; Davison et al., 2004; Susman &Evered, 1978)

1. Diagnosis(Declare Research

Framework andMethodology)

2. Action Planning

3. Intervention(Action Taking)

4. Evaluation(Assessment)

5. Reflection(Learning)

Entrance

Exit



(which also generates entries in the ac-cess logs), we filtered the entries con-taining system administrators’ usernames and Internet protocol (IP) ad-dresses. Consequently, the results of theanalysis only reflect the community mem-bers’ activities.

3. To aggregate results, we grouped theWeb pages and forms according to func-tional areas. We determined the num-ber of page views (the sum of the re-lated Web pages and forms’ page views)for each functional area and the percent-age of total page views. This percent-age is an indicator of how important thespecific functional area is for the com-munity members.

To conduct this evaluation, we usedthe Web-log analysis tool WebTrends,which can cope with the specificities ofaccess logs generated by Lotus applica-tion servers.

In the reflection (learning) phase, weanalyzed deviations between project out-

comes and expectations. The goal was togather knowledge on the relationships be-tween a community system’s design fac-tors and the community members’ activity.We discussed the differences between theoutcomes and expectations and developedhypotheses to explain them.

RESULTS

The BEC: A Learning Network

OverviewThe University of St. Gallen’s Execu-

tive MBA in Business Engineering (MBE)is a part-time postgraduate course for man-agers in leading positions. The program isintended to qualify professionals for all as-pects of business transformation (Winter,2002). At present, the MBE HSG has about90 participants in two courses and approxi-mately 270 alumni.

To support participants during thecourses and afterward, the MBE organi-zation decided to implement a community

Figure 4. BEC screenshot



system called the Business EngineeringCommunity (BEC) in 1998. By fosteringinteraction between participants, especiallyduring practice weeks and after the study,the BEC contributes to and maintains com-munity building. Because the system wasdesigned to support interaction and rela-tionship development among the currentparticipants and alumni, we focus on thesetwo groups of community members. Thecurrent release of the BEC system wasfirst implemented in February 2002.

The Design of the BECThe study of the system can be di-

vided into several phases, which are shownin Figure 5. The BEC has to support all thephases from “Introduction” to “Alumni.”Additionally, the BEC has to provide func-tions that support relationship developmentbetween participants. In the following sec-tion, each phase of the participants’ workprocess is described.

In the introduction phase, the partici-pants acquire information on the program,venues, organization, and contact persons.They edit their own profiles, view other par-ticipants’ profiles, and get used to the BEC’sfunctions. During courses, participants ob-tain information on timetables, documents,their credits, and instructors. They have todo (preparatory) exercises in groups, com-municate with one another, exchange relateddocuments, and discuss related topics. They

thereafter provide feedback on the modulesand meetings. The program also contains afour-week stay at a North-American uni-versity. Apart from information on the stayand the local companies that they will visit,the work processes in the U.S. are identicalto those in the previous phases. The thesisis written in groups of two to four persons.The participants therefore need to buildteams, and search for and discuss possibletopics. After the completion of the program,the alumni are primarily interested in main-taining the community and in networking.They organize events, search for experts,and exchange knowledge and experiences.

To support the study’s differentphases and the corresponding processes,the MBE organization decided to implementa community system. The MBE organiza-tion thus created the BEC’s functionalspecifications based on experiences withan earlier release, user feedback, and therequirements mentioned above. The newBEC system was implemented by an ex-ternal service provider and was launchedin February 2003. It is currently operatedand maintained by the MBE organization.

The BEC’s functions, which supportthe community members’ processes, canbe divided into five areas, supplemented bya sixth category for “support functions”:

• Course support: This area contains in-formation on modules and meetings, with

Figure 5. The process of study

Intro - duction

USA Thesis Alumni

Courses

Block Review Modules

Block Prepa -

ration

Business Engineering Community (BEC)



relevant documents available for down-load. Participants can also provide feed-back on modules and individual meet-ings. In addition, participants can accessstatistics relating to their study credits.To support the participants in their en-deavor to find a topic for their thesis, adiscussion forum is available in whichtopics can be suggested and discussed.

• Teamwork: In order to support the par-ticipants’ thesis work, their jointly doneexercises, and the organization of jointevents, this area enables the building ofprivate and public teams. Within a teamarea, the system provides functions suchas document exchange, collaborativedocument creation, application sharing,and a common calendar.

• Communication: The BEC offers bothsynchronous and asynchronous commu-nication. Synchronous communication issupported by a chat function or instantmessaging, video and audioconferencing, application sharing, and awhiteboard. Asynchronous communica-tion is supported by discussion forumsand an integrated e-mail system.

• Member profiles: This area containsa list of all member profiles grouped ac-cording to the courses the members at-tended. Each participant can edit his orher own profile and access other par-ticipants’ profiles.

• Content: In this area, related contentcan be published and categorized. Pos-sible content types are book recommen-dations, citations, links, attachments, andevents.

• Support functions: In addition to thefunctions just mentioned, the BEC pro-vides supporting functions such as userhelp, a feedback function, and searchand news functions.

The MBE’s cultural context in respectto technology use is primarily shaped bythe different participants. The MBE pro-gram is attended by professionals from alldisciplines and industries with differentbackgrounds and affinity for technology.Most participants have no experiences withthe community system’s technology eventhough the system’s implementation isbased on an established and widespreadIBM Lotus technology. Consequently, theMBE organization offers a short explana-tion of the community system’s functionsduring the introduction phase of the pro-gram. The functions are furthermore de-signed to be largely self-explanatory, al-though online user help is available.

The CKP-Net: An Expert Network

OverviewThe Competence Center Customer>

Knowledge>Performance (CC CKP) is aknowledge network between the Univer-sity of St. Gallen’s Institute of InformationManagement (IWI-HSG) and six majorSwiss and German financial-services com-panies that finance the competence cen-ter. The aim of this network is the develop-ment of knowledge in the areas of cus-tomer-relationship management (CRM) andknowledge management with a focus onperformance management.

Generally, the CC CKP consists of acore team — a project manager and re-searchers of the IWI-HSG — and severalof the participating companies’ employees(participants). The network has a steeringcommittee on which each participating com-pany is represented, and which meets bi-annually to discuss the network’s researchalignment.

The core team’s task is to developknowledge that the participants can usewithin their companies, while the knowl-



edge network as such should support itsparticipants’ work and relationship-devel-opment processes. To support these pro-cesses, a community system, called CKP-Net, had to be established.

The Design of CKP-NetThe CC CKP has five different pro-

cesses that have to be supported by CKP-Net: workshops, literature research, projectwork, steering-committee meetings, andrelationship building.

Workshops are conducted four timesa year and deal with varying topics. Dur-ing a workshop, the core team memberspresent state-of-the-art concepts and fu-ture trends, whereas the participants re-port on the status of related projects in theircompanies and the challenges that theycurrently face. The core team and partici-pants also take part in group work to ex-change knowledge on a new area of re-search. Both the steering-committee mem-

bers and the participants are involved inthe preparation of workshops. Eachcompany’s steering-committee member hasto select participants to attend such a work-shop. The selected participants then haveto prepare for the workshop by familiariz-ing themselves with the workshop’s top-ics. After a workshop, many of the partici-pants and steering-committee membersneed to access the presentations and re-sults of group work for utilization in theirindividual work.

Apart from the workshop documents,the participants and steering-committee mem-bers often do literature research, for example,for projects that are not carried out in col-laboration with the competence center. Con-sequently, CKP-Net has to support the publi-cation of research documents and a topic-oriented structured search as well.

The core team members also supportthe participating companies’ employees bymeans of projects within their respective

Figure 6. CKP-Net screenshot



companies. Project work that the partici-pants and core team carry out requires anintensive exchange of documents, for ex-ample, documents explaining the enterpriseenvironment, project plans, and concepts.CKP-Net therefore has to facilitate theexchange of documents and the collabora-tive creation of documents.

It must also be possible to publishagendas, presentations, and the results ofdiscussions in steering-committee meetingson CKP-Net to allow the steering-commit-tee members easy access to these.

In order to build relationships withinthe competence center, it is useful for thecore team, participants, and steering com-mittee to communicate with one another attimes other than during the workshop meet-ings. All stakeholders’ contact details shouldtherefore be published within CKP-Net.

To support all these processes, thecore team decided to implement a commu-nity system. Because previous competencecenters had already worked with commu-nity systems for several years, CKP-Netwas designed based on experiences withthese systems and the above-mentionedrequirements. In January 2003, it waslaunched and had an expected life span oftwo years. Its functions, which support thecommunity members’ processes, can bedivided into the following areas:

• General Information: This tab con-tains documents on project plans, a listof publications, and important links.

• Team Information: Two documentsprovide the contact information of coreteam members, participants, and steer-ing-committee members, as well as pho-tos of the core team members.

• Workshops: This tab includes docu-ments pertaining to planned and con-ducted workshops (information on loca-

tion, agenda, workshop presentations,and photos of conducted workshops).

• Research Topics: Included under thistab are several academic papers and pre-sentations dealing with research topicsthat are relevant to the participants andsteering-committee members.

• Project Rooms: Each of the partici-pating companies has its own collabora-tive work space that can be accessedvia this tab. A work space includes sup-port for the exchange and collaborativecreation of documents, as well as dis-cussion forums.

• Steering Committee: This tab containsdocuments about planned and conductedsteering-committee meetings (informa-tion on location, agenda, presentations,and photos of conducted steering-com-mittee meetings).

• Archive: Integrated under this tab areworkshop documents and the documen-tation of previous competence centers’specific research topics.

• Chat: By clicking on this button, eachCKP-Net user can activate an inte-grated Lotus Sametime client applica-tion for awareness and instant messag-ing (AIM). This client application alsoprovides opportunities for audio andvideo meetings, as well as applicationsharing.

• Support Functions: Further supportfunctions include integrated help, a feed-back function, and search and newsfunctions.

Summary of ResultsFigure 7 summarizes the previous

sections’ results, as well as the activityanalysis’ results (rows labeled “Process-Supporting Functions & Use of Functionsby Community Members”). The first col-umn in this row contains the process-sup-



porting functions and functional areas de-scribed in the previous section. The sec-ond column (“Page Views”) indicates howoften a document related to the specificfunction was viewed. For example, the fig-ure 6,822 in the first row indicates thatdocuments related to information aboutmodules and meetings (e.g., timetables, sta-tistics on own credits, course presentations)were accessed 6,822 times by the commu-nity members. The third column shows thepercentage of page views in relation to thetotal number of page views (42,745).

Comparison of the Research SitesTo compare the research sites, we

can use Figallo’s (1998) classification cri-teria for communities, which is based oncommunity members’ behavior (e.g., de-gree of personal interactivity, subject scope,cohesion of members), complemented withthe research model’s criteria, except forusability and satisfaction (because thesewere not measured but approximated bymeans of activity measurement).

• Degree of personal interactivity:Both communities have a high degreeof face-to-face interactivity during per-sonal meetings (courses and work-shops). Interaction by means of the com-munity systems is therefore only mod-erate and confined to the absolutely nec-essary.

• Subject scope: The subject scope inboth communities is relatively narrow asboth of them deal with a specific sub-ject area. In the BEC’s case, this is busi-ness engineering, while in CKP-Net’scase, this is CRM and KM.

• Cohesion of members: In respect toboth research sites, the cohesion of mem-bers is sustained by a semiformal organi-zation: in the BEC’s case, the course or-

ganization, and in the CKP-Net’s case, theorganization of the competence center.

• General cultural Context: Both com-munities are interorganizational andtherefore have members from organi-zations with different cultural contextsand affinities for technology. In theBEC’s case, the members’ cultural con-text is highly heterogeneous, whereas inCKP-Net’s case, the cultural context ismore homogeneous because memberscome from the same industry and busi-ness departments.

• Work and relationship-developmentprocesses: The two communities’ workprocesses reveal similar characteristics.There are processes for the preparationand wrap-up of person-to-person meet-ings (courses and workshops), and col-laborative processes to create docu-ments or project outcomes (teamworkand project collaboration). Moreover,both communities demonstrate the needfor relationship development. However,the MBE community has a greater needfor the support of communication pro-cesses (discussions) than CC CKP.

• Technology/media choice: Both com-munities employ similar technology fortheir community systems.

• Experience and training: Becausecommunity members have no previousexperience with the chosen technology,both communities offer their membersa short explanation of the system func-tions in an introductory person-to-per-son meeting as well as online help withinthe systems. The systems’ functions alsohave to be widely self-explanatory toavoid the need for extensive training.

In conclusion, both communities showvery similar characteristics. This strength-ens the generalizability of our findings.



Figure 7. Summary of results

BEC CKP-Net

Objective Qualify professionals for all aspects of business transformation

Interorganizational knowledge development in the areas of CRM and KM

Members • Current participants of the MBE HSG

(ca 90) • Alumni of the MBE HSG (ca 270)

• Steering committee (ca 7) • Participating companies’ employees (ca 50) • Core team members (ca 7)

General Cultural Context Regarding Technology Use

Professionals from all disciplines and industries with different backgrounds and technology affinity

Financial industry professionals from marketing, sales, service, and IT departments with different backgrounds and technology affinity C

omm

unity

Work Processes and Relationship Development

Participants and alumni: • Access course information • Communicate • Work together • Search for/edit profiles • Exchange information/documents • Organize events/meetings Participants: • Access credits • Prepare thesis • Give feedback

Participants and steering committee: • Project collaboration • Workshop preparation • Workshop document retrieval • Document search • Communication Steering committee: • Steering-committee meeting document retrieval

Goal of the Community System

Support work processes and relationship development among current participants and among alumni

Support work processes and relationship development among core team members and participants

Technology Integrated solution using Lotus Notes/Domino, Lotus Team Workplace, Lotus Sametime

Integrated solution using Lotus Team Workplace and Lotus Sametime

Page Views

%

Course Support

Page Views

%

Information about modules and meetings

6,822 15.96 General information 36 3.62

Feedback on modules 1,453 3.40 Member profiles 154 15.48

Feedback on meetings 150 0.35 Workshop information 620 62.31

View credits 22 0.05 Research documents 63 6.33

Thesis topic bourse 4,231 9.90 Project collaboration rooms 40 4.02

Finished thesis topics 4,726 11.06 Steering committee 36 3.62

Teamwork Archive 46 4.62

Teamwork area 5,968 13.96 Chat 0 0.00

Communication Support functions 0 0.00

Chat 436 1.02

General discussion 1,784 4.17

Course discussion 5,909 13.82

Member Profiles

Member profiles 7,022 16.43

Content Administration

Content 1,317 3.08

Browse 2,905 6.80

Process-Supporting Functions & Use of Functions by Community Members (activity) [February to May 2003]

� 42,745 100 � 995 100

Experience Most members had no experiences with the community system or its technology before using the system.

Most members had no experiences with the community system or its technology before using the system.

Com

munity System

Training

• Short explanation of functions during introductory phase of the program

• Functions are widely self-explanatory • Online user help

• Short explanation of community system’s functions during workshops (30 minutes)

• Functions are widely self-explanatory • Online user help



DISCUSSIONIn this section, we present the results

of the action research process’s reflection(learning) phase (Figure 3). We discuss theresults of the previous section in order todeduce propositions for an improved de-sign of community systems that supportprofessional communities.

The functional areas of both commu-nity systems were designed in accordancewith the community members’ work andrelationship-development processes. How-ever, there are significant differences re-garding the community members’ activityin respect to the use of the functions. Inthe following sections, we discuss the im-plications for the support of work processesand relationship development.

Support of Work ProcessesIt is apparent that the most frequently

visited functional area in CKP-Net was“Workshop Information,” and in the BEC’scase, this was “Course Support.” In con-trast, the least visited area in the BEC was“Content Administration,” and “SupportFunctions” and “Chat” in CKP-Net. Gen-erally, we observe that functions related toface-to-face meetings — the preparationfor or wrap-up of meetings — such as“Course Support,” “Teamwork,” and“Workshop Information” are used far morethan other functions. This cannot be attrib-uted to a difference in experience or train-ing because these are roughly the same forthe different functional areas. Our conclu-sion is that in these professional communi-ties, interaction by means of face-to-facemeetings plays a primary role, whereas in-teraction by means of the community sys-tems plays a secondary role. These sys-tems are therefore only regarded as sup-portive tools for face-to-face meetings andnot as a primary means of communicationas they are in virtual communities.

Moreover, we can see that functionsthat are directly connected to work pro-cesses, for example, information aboutmodules and meetings and “Course Dis-cussion” in the BEC, or “Workshop Infor-mation” in CKP-Net, are used much morefrequently than general functions like“Archive,” “Chat,” or “General Informa-tion” in CKP-Net.

The functions of community systemssupporting professional communities shouldtherefore be designed so that they first sup-port the community members’ face-to-facemeetings, and second, so that they are di-rectly related to a specific work process.Community members need to know howthey can directly benefit from a specificfunction in their work processes. If a func-tion is not explicitly assigned to a work pro-cess, community members do not realizeits usefulness.

Support of RelationshipDevelopment

An interesting observation was thatin both community systems, the “MemberProfiles” area was the second most fre-quently accessed area. Most of the com-munity members meet on a regular basis,either in workshops or in courses, and cantherefore engage in relationship buildingduring these events. While research hasconfirmed that face-to-face communicationoffers a better means for relationship build-ing than CMC (Chidambaram, 1996), the“Member Profiles” areas are neverthelessextensively used. Relationship-developmentsupport therefore seems to be an integralfunctional part of community systems sup-porting professional communities.

In respect to the BEC, the “MemberProfiles” activity can be partly explained bythe fact that the majority of the communitymembers are alumni who do not participatein courses, but are only interested in relation-



ship development and maintenance. More-over, participants in different courses — whodo not regularly meet one another — mayuse the system to retrieve additional informa-tion about people whom they have met.

In CKP-Net, the “Member Profiles”area seems to complement regular face-to-face communication in workshops verywell. Members may use it to retrieve addi-tional information about people whom theyhave met in workshops, or to retrieve con-tact details so that they can make telephonecontact.

On the other hand, functions that canpotentially be used for relationship-develop-ment processes, such as discussion forumsor chat, are not used very much. This leadsus to the conclusion that the “Member Pro-files” area is essentially being used by com-munity members to obtain detailed personalinformation on other community members,for example, information on their companies,job descriptions, and work areas, to extendtheir knowledge on fellow community mem-bers in order to facilitate (or enable) non-computer-based communication.

Besides the functions that supportwork processes related to face-to-facemeetings and relationship development, theothers are seldom used. This is particularlytrue of most synchronous and asynchro-nous communication functions; for example,the “Chat” function in CKP-Net was neverused and the project rooms were rarelyused. In the BEC, “Chat” and “GeneralDiscussion,” too, were rarely used, while“Teamwork Area” and “Course Discus-sion” were by contrast used quite fre-quently, mainly for the exchange of docu-ments and the discussion of topics relatedto courses.

This supports our hypothesis that theprimary means of communication betweencommunity members is face-to-face meet-ings. The community systems are only used

for interaction among community membersif the functions offer support for face-to-face meetings as in the case of “TeamworkArea” and “Course Discussion.” Moreover,communication functions are not used forrelationship development, but for the sup-port of work processes. Therefore, com-munication functions that do not directly sup-port work processes, like, for example, the“Chat” functions, are obviously dispensable.

CONCLUSIONAND OUTLOOK

Our goal was to analyze which de-sign factors of community systems influ-ence community activity, and how commu-nity systems supporting professional com-munities should therefore be designed.Based on the findings of a study of two real-world professional communities — a learn-ing network and an expert network — wedrew a few conclusions (Figure 8) that arevalid for the communities analyzed. Theseconclusions may help in the design of com-munity systems for professional communi-ties. However, a study with a statisticallyrelevant sample must be performed beforeour conclusions can be assumed as gener-ally valid for all professional communities.

In communities that employ a mix ofcommunication modes of CMC, telephone,and face-to-face communication, the com-petition between the different media has tobe considered in supporting community sys-tems’ design in addition to other factorsrelevant to the design of community sys-tems (Figure 8). Our study shows thatface-to-face communication is the primarymode of communication; community sys-tems only have a supporting function. Thisleads us to deduce some design guidelinesfor systems supporting such communities.

Generally, community systems haveto support professional communities’ work



processes and relationship-developmentprocesses. Important functions for work-process support are functions that supportface-to-face meetings (for the preparationfor and wrap-up of meetings) and that ex-plicitly support specific work processes. Incontrast, functions that do not support face-to-face meetings or are not explicitly as-signed to work processes are dispensableand should not be implemented. Importantfunctions for relationship development arefunctions that enable or facilitate face-to-face meetings, for example, member pro-files that enable members to retrieve fur-ther information on other community mem-bers’ companies, job descriptions, work ar-eas, contact details, and so forth. On theother hand, functions for general computer-mediated communication, for example, chator general discussion areas, are dispensable.

Critically reviewing our work, it waspossible to draw conclusions for improvedcommunity-supporting systems design, butfurther empirical research, for example, bycarrying out qualitative interviews withcommunity members, is necessary to in-terpret our findings and to learn what mo-tivates the community members’ virtualactivity.

One important question remaining tobe answered is what influence the design

of the functions in a community system hason a professional community’s overall per-formance. To answer this question, wehave to analyze and combine measurementresults from different performance levels(activities, output, and value; see Figure 2).Currently, we are working on metrics tomeasure performance in terms of outputand value in the analyzed communities. Ourresearch objective is to show the link be-tween the technology application and busi-ness results of a professional community.

REFERENCESBach, V., Vogler, P., & Österle, H. (Eds.).

(1999). Business Knowledge Manage-ment: Praxiserfahrungen mit Intranet-Basierten Lösungen. Berlin, Germany:Springer.

Bieber, M. et al. (2002a). Toward virtualcommunity knowledge evolution. Jour-nal of Management Information Sys-tems, 18(4), 11-35.

Bieber, M. et al. (2002b). Towards knowledge-sharing and learning in virtual professionalcommunities. In Paper presented at the35th Annual Hawaii InternationalConference on System Sciences, Ha-waii.

Checkland, P., & Holwell, S. (1998). Ac-tion research: Its nature and validity.

Figure 8. Design guidelines for community-supporting systems in professional communities

Important Functions Unimportant Functions

Relationship Development

Functions that enable or facilitate face-to-face meetings, for example: • Member profiles to retrieve further

information on the company, job description, work areas, contact details, and so forth, of community members

Functions for general computer-mediated communication, for example: • Chat and • General discussion

Process support Work Processes

Functions that: • Support face-to-face meetings

(preparation, wrap-up) and • Are directly assigned to work

processes

Functions that: • Do not support face-to-face meetings or • Are not directly assigned to work processes

(such as general computer-mediated communication)



Systemic Practice and Action Re-search, 11(1), 9-21.

Chidambaram, L. (1996). Relational devel-opment in computer-supported groups.MIS Quarterly, 143-165.

Davison, R. M., Martinsons, M. G., &Kock, N. (2004). Principles of canoni-cal action research. Information Sys-tems Journal, 14(1), 65-86.

Etzioni, A., & Etzioni, O. (1999). Face-to-face and computer-mediated communi-ties: A comparative analysis. The Infor-mation Society, 15, 241-248.

Figallo, C. (1998). Hosting Web commu-nities: Building relationships, increas-ing customer loyalty, and maintain-ing a competitive edge. New York: JohnWiley & Sons.

Godio, C. (2000). Building a virtual profes-sional community: The case ofPoolweb.it. In Paper presented at theThird International Conference on Vir-tual Communities, London.

Hagel, J., & Armstrong, A. G. (1997). Netgain: Expanding markets through vir-tual communities. Boston: HarvardBusiness School Press.

Hillery, G. A. (1955). Definitions of com-munity: Areas of agreement. Rural So-ciology, 20, 118-120.

Hiltz, S. R., & Johnson, K. (1990). Usersatisfaction with computer-mediatedcommunication systems. ManagementScience, 36(6), 739-764.

Kraemer, W., & Müller, M. (1999). Virtualcorporate university: Executive educationarchitecture and knowledge manage-ment. In A.-W. Scheer (Ed.), Electronicbusiness and knowledge management.Heidelberg, Germany: Physica.

Mansell, G. (1991). Action research in in-formation systems development. Jour-nal of Information Systems, 1, 29-40.

Markus, U. (2002). Integration derVirtuellen Community in Das Cus-

tomer Relationship Management. Un-published dissertation, University ofSaarland, Saarbrücken.

McDermott, R. (2002). Measuring the im-pact of communities. Knowledge Man-agement Review, 5(2), 26-29.

Poplin, D. E. (1979). Communities: A sur-vey of theories and methods of re-search (2nd ed.). New York: McMillan.

Rheingold, H. (1998). Virtual communities.In F. Hesselbein (Ed.), The communityof the future. San Francisco: Jossey-Bass.

Schulte, T. (1999). Group computingworkspace. Unpublished dissertation,University of St. Gallen, St. Gallen.

Stanoevska-Slabeva, K., & Schmid, B. F.(2000). Requirements analysis for com-munity supporting platforms based on themedia reference model. ElectronicMarkets, 10(4), 250-257.

Susman, G. I., & Evered, R. D. (1978). Anassessment of the scientific merits ofaction research. Administrative Sci-ence Quarterly, 23(14), 582-603.

Sutton, W. A., & Munson, T. (1976). Defi-nitions of community. New York:American Sociological Association.

Watson-Manheim, M. B., & Belanger, F.(2002). Exploring communication-basedwork processes in virtual work environ-ments. In Paper presented at the 35thAnnual Hawaii International Confer-ence on System Sciences, Hawaii.

Wenger, E. (1997). Communities of prac-tice: Learning, meaning, and identity.Cambridge: Cambridge UniversityPress.

Wenger, E., McDermott, R., & Snyder, W.M. (2002). Cultivating communities ofpractice: A guide to managing knowl-edge. Boston: Harvard Business SchoolPress.

Wiesenfeld, B. M., Raghuram, S., & Garud,R. (1999). Communication patterns as



determinants of organizational identifi-cation in a virtual organization. Organi-zation Science, 10(6), 777-790.

Malte Geib has been a researcher and PhD student at the Institute of Information Management,University of St. Gallen, Switzerland since 2001. His research focuses on customer relationshipmanagement and customer knowledge transfer in networked enterprises, especially in thefinancial services industry. He is currently (2005) a visiting researcher at the Center for DigitalStrategies, Tuck School of Business at Dartmouth College, USA. He earned a master’s of sciencein information systems from the University of Muenster, Germany, and has studied at theQueensland University of Technology, Australia. He has worked as a software engineer for IBMGermany and as a consultant for several Swiss and German companies, including DeutscheTelekom AG, Credit Suisse, and PostFinance.

Christian Braun has been a researcher and PhD student at the Institute of InformationManagement, University of St. Gallen, Switzerland since 2002. His research interests are methodengineering, reference modeling, outsourcing, and object-oriented software development, aswell as the design of information systems, especially in the financial services industry. From1997 to 2002, he studied computer science and business administration at the University ofMunich, Germany. He received a diploma in computer science in 2002. He is currently jointlyresponsible for the execution of an executive MBA program at the University of St. Gallen.

Lutz Kolbe has headed the Competence Center Customer Knowledge Performance (CC CKP)and taught at AACSB-accredited University of St. Gallen, Switzerland since July 2002. Hisresearch interests are customer relationship management, security management, and advancedtechnologies in the residential environment. After having worked as a financial consultant, hestudied information management at Brunswick Technical University, Germany, where he receivedhis master’s. He continued working on his dissertation at Freiberg Technical University, Germany,and the University of Rhode Island, USA and received his PhD in 1997. After that he worked atDeutsche Bank in Frankfurt and New York where he became managing director in 2001.

Prof. Dr. oec. Walter Brenner is professor of information systems at the University of St. Gallenand managing director of the Institute of Information Management. After earning his graduateand doctorate degrees and his venia legendi from the University of St. Gallen, he workedseveral years for the Alusuisse-Lonza AG based in Basel, Switzerland as the head of applicationdevelopment, among other positions. From 1993 until 1999 he was professor for businessadministration and information management at the TU Bergakademie Freiberg, Germany andsubsequently professor for information management and business administration at theUniversity of Essen, Germany. His research focus is on integrated information management,customer relationship management, and innovative technologies.

Winter, R. (2002). An executive MBA pro-gram in business engineering: A curricu-lum focusing on change. Journal of ITEducation, 1(4), 279-288.

toward improved community- supporting systems design

Documents