GLOBAL COMMUNICATION

Systems Development by Virtual Project Teams 447 Copyright © 2003, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited. Copyright Idea Group Inc. Copyright Idea Group Inc.

A Comparative Study of Four Cases David Croasdell Washington State University, USA Andrea Fox Washington State University, USA Suprateek Sarker Washington State University, USA EXECUTIVE SUMMARY Business organizations and global partners are increasing their utilization of virtual project teams to enhance competitive advantages in the global market. More than ever, organizations are using virtual teamwork to bridge time zones and geographic distances. The use of virtual work environments has spurred interest in understanding how team members interact and collaborate over the life of a project. Not surprisingly, organizations are trying to understand what factors are determinants of success with respect to virtual teams.

Increasing network bandwidth, continuously improving communication technologies, shift- ing global economies, and changes in social practices have caused business managers to reconsider traditional practices. This paper provides a comparative case study of four cross- cultural virtual project teams as they analyze, design, and develop information systems. BACKGROUND Teams are an integral part of organizational life (Gersick, 1988). Recent trends in globalization and advances in telecommunications technologies have enabled the use of distributed teams, especially those involved in Information Systems Development (ISD).

These “virtual teams” consist of geographically dispersed team-members who interact using 701 E. Chocolate Avenue, Hershey PA 17033-1240, USA Tel: 717/533-8845; Fax 717/533-8661; URL-http://www.idea-group.com IDEA GROUP PUBLISHING This chapter appears in the book, Annals of Cases on Information Technology, Volume 5, edited by Mehdi Kosrow-Pour. Copyright © 2003, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited. 448 Croasdell, Fox & Sarker Copyright © 2003, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited. Copyright Idea Group Inc.

Copyright Idea Group Inc.Copyright Idea Group Inc. information and communication technologies (ICTs) such as email, groupware, video, and computer-based conferencing systems. Briggs, Nunamaker, and Sprague (1998) have ob- served that while the demand for virtual teams grow, “little is known on how to actually conduct team telework” (p. 11). This case study provides some insights into virtual teamwork, specifically in the context of ISD. Virtual Project Teams Work teams described in this comparative case study were asked to identify the business problem associated with their project, develop a plan for addressing the problem using an information system to enable the solution, analyze design alternatives, define system requirements, and develop a working solution to address the business problem. Four project teams were observed as they worked to identify, initiate, plan, design, develop and implement information systems using formal system development methodologies. Project teams included some members that were co-located and some members that lived and worked halfway across the globe.

Teams coordinated their efforts in a virtual environment using WebCT™ a product designed to provide support for virtual collaborative computing environments by enabling synchronous and asynchronous communication between local and remote users. WebCT™ includes mechanisms for online discussion forums, synchronous chat, file transfer, and shared calendaring. The application was the primary tool used to support the communication, collaboration, and coordination among team members in each virtual team. Each team was given training to familiarize team members with tool specific functions. Project team members were experienced information technology users and proved quite proficient in their use of the tools. Individual team members posted discussion topics to a local team folder or to a global area accessible to the entire team. Discussion pages provided an environment for team members to read and reply to discussion topics asynchronously. Chat sessions provided synchronous communication in which multiple participants exchanged thoughts using typed dialog. Participants in chat sessions could view messages sent by other participants in “real- time.” With file sharing, team members posted project documents to a shared space.

Documents placed in the common project directory were reviewed, updated, and re-posted as necessary. The calendar tool gave teams the ability to schedule events to a master team calendar for all members to see. Facilitators observed development efforts, provided guidance and arbitrated team conflicts. Table 1 shows the extent to which each team used the tools available in WebCT™ over the period of the projects. Team A held the most chat sessions (16), Team B scheduled far more events than other groups (26), Team C had a greater propensity for sharing files (68), and all of the groups contributed several discussion messages. Table 1. Use of WebCT™ Tools in Virtual Teams Team Chat Session s Calenda r EventsFiles Share d Discussio n Message s A 16 1 29 184 B 10 26 38 308 C 11 9 68 318 D 13 8 33 366 Table 1. Use of WebCT™ Tools in Virtual Teams Systems Development by Virtual Project Teams 449 Copyright © 2003, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited. Copyright Idea Group Inc. Copyright Idea Group Inc.

Project A Aerial Painters Incorporated (API) is a private company located in the Pacific Northwest region of the United States. The company employs approximately 65 workers. API’s business is painting medium to large-sized commercial airplanes. The primary contact for this project was Bruce Drago, API’s business manager. Mr. Drago wanted to develop a Web presence incorporating the marketing goals of API. The plan for the project included developing a system that provided file-sharing capabilities, networked security cameras, an e-commerce Web site and a redesigned user interface. Bruce wanted the new system to integrate the companies existing marketing plan and an employee-tracking database. He requested changes that would convey corporate purpose, competitive position and industry presence.

Bruce contacted Team A to discuss his project. The team was composed of five members from Norway and five from the U.S.

API has made a significant investment in building the image of the company. They wanted their Web presence to reflect an image of high quality and high customer satisfaction.

Mr. Drago wanted a Web site to offer information about API’s mission, goals, and products and services. In addition, the site was to include a Web-enabled video clip providing an introduction to API’s operations. Mr. Drago wanted employees to have access to updated information, company news and announcements real time. Potential employees could use the pages to investigate employment opportunities and apply for positions within the company.

An e-commerce element would allow API to offer stenciling services to its customers.

Bruce wanted a portion of the Web site to be dedicated to an Intranet allowing API management access to networked cameras and private employee information. The ability to include networked cameras was an important design feature for API managers. Past accidents in API’s hangers resulted in injuries to workers as well as damage to customer planes. Network cameras would allow managers to review accidents and stored video footage to help in accident analysis and implementation of new quality controls. The system would give managers the ability to monitor work areas from anywhere.

API’s information systems have historically been paper-based. Physical artifacts created high overhead costs due to the expense of long distance phone services, faxed messages and Federal Express shipping charges. Work orders followed an interactive development process using surface based snail mail to pass designs and modifications between clients and the company. The new system was to provide a document transfer capability between engineers and clients. Online chat and discussion forums were to facilitate communication and shorten product design and development cycles.

Team A was asked to design and build a Web presence based on the guidelines formulated by Bruce Drago. Beyond creating functional pages, the team had to address issues of security and liability, increased communication with business partners, dynamic marketing initiatives and interfaces with the companies human resources department. An iterative development process was initiated between API and project team members. Throughout these early stages of the project, Team A successfully utilized communication tools to develop a strong social connection among team members. In part, this social connection helped the team deliver quality products to their client. Project B The sports and recreation division at State U. provides recreational activities in an organized and safe manner for a community of 21,000 people. Historically, the division used 450 Croasdell, Fox & Sarker Copyright © 2003, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited. Copyright Idea Group Inc.

Copyright Idea Group Inc.Copyright Idea Group Inc. an informal, paper-based information system. Data collected from various typed and handwritten forms was used to track event information, employee data and work shift assignments. File cabinets were used to store information related to workers, activities and assigned tasks. For example, referees and umpires were regularly scheduled to officiate recreational sports activities. Using a paper system, division administrators sorted through hundreds of forms to determine the availability of officials and create weekly officiating schedules. The mechanism was complicated by the fact that multiple activities—requiring different officiating experience and background—occurred at multiple venues simulta- neously. It was often necessary to modify schedules to reflect substitutions and resched- uling when officials were unable to attend an event.

A paper-based system limited the ability of administrative personnel to schedule officials in an efficient and effective manner. A centralized data store (file cabinet) did not allow multiple users to access the information concurrently. The time and effort required to maintain the systems was a source of frustration for division personnel. System administra- tors requested a new, automated information system to manage event scheduling. The purpose of the system was to coordinate sporting events, recreational activities and scheduling of required workers. Users requested the system be built using Microsoft Access as a backend database with a Web-style browser interface. The system was to be accessed anywhere an Internet connection and Web browser were available. Remote access would provide relevant information to individuals in the community looking up event information.

Online access would allow officials to check schedules, update records and trade work shifts online. Finally, a Web-based system would allow administrators to query the database for relevant event and personnel information.

A project team comprised of five members each from the United States and Norway were asked to analyze the needs of the sports and recreation division. Their task was to design and implement an information system that would address the needs of the division. The team successfully completed the project and provided a working system to the sports and recreation division, but the team did not forge the social interactions necessary to bind members of the team. At the midway point of the project, a crisis in the relationship among the U.S. members and the Norwegian members was narrowly averted. The lack of social glue in the group nearly caused the project to breakdown. The team faced many challenges that were not directly related to project tasks. Lack of any real social interaction created several obstacles that the group had to overcome in addition to completing the technical require- ments for the project.

Project C Envision is a small innovation company in Norway. The company was established in February 2000. Sixteen employees develop products for use on the World Wide Web. Rolph Lyndgren is president and CEO of the company. Mr. Lyndgren requested an information system that would serve as a knowledge base in support of Envision’s primary business function: the development of innovative products to enable e-business solutions. Employ- ees of the company conduct market research via the Internet. Information from relevant Web sites is stored using tools and features found in most Web browser software (e.g., Favorites folders). Web browsers used by Envision employees provide the ability to bookmark URL addresses for later reference. However, bookmark information and Web page content was not available outside individual client computers unless users actively posted the URL’s to a company server or e-mailed the information to interested co-workers. Mr. Lyndgren envi- Systems Development by Virtual Project Teams 451 Copyright © 2003, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited. Copyright Idea Group Inc. Copyright Idea Group Inc.

sioned a “global bookmarks” folder that would serve as a knowledge base within the company. As employees identified relevant links on the Web, they could store the links in a centralized bookmarks file. Such a system would allow the company to stay abreast of co- workers efforts while also improving efficiency and reducing overlap. Ultimately, Mr.

Lyndgren wanted centralized storage and access for all company documents.

Because Envision is an innovation company, they are quite concerned with security.

Any centralized information system would require mechanisms for physical security as well as user login and authentication measures. In addition, files stored on centralized servers would require safeguards against outside intrusion. The system would also require a configuration management component that would notify employees of file updates and establish access privileges for company employees.

As with the other projects, Team C was comprised of ten team-members, five each from the United States and Norway. The team’s objective was to develop a knowledge base to share URLs and track useful Web content. The team was successful in creating a functioning system that met the specifications of the client. Team members established strong social bonds throughout the project. In large part, this social cohesion was a driving factor behind the project. Team C consistently utilized all communication and collaboration tools made available. However, the most prevalent form of communication was virtual chats. Sessions were held at least once each week and very often two or three times a week. Almost every team member attended each chat session. Frequent communication created a strong social bond and allowed the team-members to grow quite close. As the project developed, the team relied heavily on the social relationship to overcome difficulties with system development.

Project D Nembus is an independent Norwegian company established in 1933. The company has offices in Norway, the United Kingdom, Germany, Italy, the United States, Korea, China, and Taiwan. Offices worldwide employ more than 400 workers in 11 countries. Engineers comprise the majority of the workforce. Nembus’s primary business function is to provide testing, inspection and certification services for electrical equipment, machinery and systems.

Certification is necessary for electrical products to be released to the market. Manufactures would like certification to take place quickly and efficiently so that they can gain access to their markets as quickly as possible. Engineers utilize technologically advanced testing laboratories throughout the world to test electrical components and products. There are six testing laboratories in Europe, two in the United States and three in the Far East.

Nembus planned to replace its current information systems with Internet enabled applications to create broader access and synchronization for their employees across the globe. The information system in use required engineers to complete a variety of paper forms to report test results. Initially, corporate secretaries transcribed over 1,000 different forms and reports into digital format using Microsoft Word. Eventually, supervisors encouraged engineers to complete the certification forms themselves thereby circumventing the need for the army of corporate secretaries. However, the goal to reduce overhead and increase efficiency had some unintended consequences.

Test engineers at Nembus work autonomously in test labs around the world. As such, they are required to work with different criteria and generate different reports depending on local, cultural and political regulations. Inconsistencies in forms and reports occurred due to differences in language, terminology, and legal statutes in addition to individual differ- ences in proficiency with the application software. Over time many of the documents were 452 Croasdell, Fox & Sarker Copyright © 2003, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited. Copyright Idea Group Inc.

Copyright Idea Group Inc.Copyright Idea Group Inc. created and modified in an ad hoc fashion. The nature of the process produced an environment in which forms and reports were often incorrect, incomplete, and/or redundant.

Engineers spent considerable time creating document templates in addition to filling them in.

The net effect was a dramatic decrease in the efficiency of the testing labs (not to mention potentially perilous certifications). Managers at Nembus recognized they had a problem.

They resolved to create a new information system that would reestablish a consistent look and feel to the certification documents and allow users to input, store, and search electronics testing forms and reports.

The new system was to be Internet-enabled and provide a repository of previously completed forms that could be retrieved and modified as appropriate for product certifica- tions. Engineers would be afforded the ability to retrieve previously used reports and update only relevant information. The new system would increase consistency and efficiency across the organization by serving as a knowledge base to disseminate testing guidelines and results across the organizations where previously the information was held in the local testing facility. Test report forms (TRF) and a fourth generation relational database (SMOKK) were integrated to provide information about the products Nembus tested and the customers who used Nembus’s services.

Team D was comprised of four members from the Unites States and five members from Norway. The team’s objective was to develop a prototype for Nembus engineers. While team D completed the project and provided a system to the client, they also demonstrated many characteristics of an unsuccessful team. To call the completion of the project and delivery of the prototype a success was to stretch the truth. The final deliverable was mediocre at best.

The system may not have been delivered at all if not for the major efforts of a single U.S. team member. The team lacked a strong presence of both social communication and task-related communication.

SETTING THE STAGE One useful way for evaluating virtual teams is to examine the quantity and frequency of communication among the members. Literature suggests that two dimensions in the communication are important: the social dimension and the task dimension. Social dimension is used to describe communication that focuses on building social relationship within the virtual team. Aspects of social dimension include shared understanding, mutual trust and social bonding among team members. Task dimension refers to communication that focuses on specific tasks that affect project completion. Aspects of task dimension include a common project goal, domain related knowledge and skills, task coordination, and modes of interac- tions within the team. Social Dimension High Low High Ideal Professiona l Task Dimension Low Social Bein g Apathetic Figure 1: Virtual Team Archetypes Figure 1. Virtual Team Archetypes Systems Development by Virtual Project Teams 453 Copyright © 2003, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited. Copyright Idea Group Inc. Copyright Idea Group Inc.

Copyright Idea Group Inc.Copyright Idea Group Inc. Lau et al. (2000) describe four archetypes of virtual teams based on the communication patterns and communication content. The level of social and task dimensions evident in the team’s interactions describe each archetype (Figure 1). Common actions are defined for each archetype and can be used to determine which archetype a virtual team is placed in. All virtual team communications exhibit various levels of social communication and/or task communi- cation.

Teams described in this manuscript can be evaluated using information about the various methods of communication presented in Figure 1. Common behaviors of virtual teams are presented in Table 2. The WebCT™ tools used to enable the virtual team environments were used to varying degrees by each of the project teams. The extent to which each team used the tools and the way in which team members utilized the tools seemed to trigger social or task orientations in the groups. The following sections illustrate team interactions and show how the teams went about their work along the orientations presented in Figure 1. CASE DESCRIPTION Phase I: Initiation and Planning Organizations identified for the virtual team projects had a defined set of requirements for the information systems they envisioned. The virtual work teams assigned to create the systems conducted thorough analysis before embarking on development. Each team wrote a narrative description of the project in order to clarify scope and direction. Teams also specified resource requirements and assessed the feasibility of the project. Finally, each team developed a baseline project plan to manage the progress of the work. The following passages provide information regarding team interactions throughout the initiation and planning stage of system development. Social Dimensi on Hi g h Low Time - Space Hi g h IDEAL TEAMS Teams members understand each other’s norms, values and experi ences irrespective of location. Team members trust and rely on each other completely. They share humor, gestures and personal st ories together. They often have common domai n knowledge and skills leading to easy interactions with each other through different technologies. Ideal teams are we ll coordinat ed i n their project tasks, deli verables and timel ines. Teams members are sensitive to each other’s time and space differences for reasons of projec t coordinat ion and personal respect. Teams progress rapidly to the mutual communication stage to build social relati onship and work cohesively on projects. PROFESSIONAL TEAMS These teams have a common overall project goal. Team members possess the necessary domain knowledge and ski ll s requi red to do th e job. They are well coordinated with regard to tasks, in their project tasks, deliverables and ti mel ines Team members engage in substantiv e interactions in terms of intensity and meaningfulness. These teams fluctuate between bi-directional and mutual communication when working on project task s and deliverables. Team members spend little time trying to understand each other’s norms, values and experiences. They typically make little effort to share humor or personal stories. Table 2a. Behaviors of Virtual Teams According to Archetype (Based on Lau et al., 2000) 454 Croasdell, Fox & Sarker Copyright © 2003, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited. Copyright Idea Group Inc.

Copyright Idea Group Inc.Copyright Idea Group Inc. Team A—Aerial Painters, Inc. Initial communication between team members occurred in the form of a discussion posting that provided names and contact information for the Norwegian team members.

Along with contact information came a promise that members from Norway would “post some lines about [themselves] individually” and a request that “it would be nice if [U.S. members] could do the same.” Eight out of ten of the team members quickly provided information about themselves ranging from topics of age, personal interests, favorite movies or music, and relevant skills and experience. A few members included links to personal Web sites or attached photos of themselves. Most of these initial postings concluded with the team members expressing their excitement about the project and enthusiasm about working with the various members (e.g., “I look forward to working with all of you. I feel confident that we can learn from each other!”) Less than 48 hours after the first discussion posting was made, the team began sharing information regarding the project. Members in the United States initially took the lead in this area, as they had direct access to client. All documentation received from the client was posted in shared file sections to inform all members of project communication. This practice proved especially valuable because team members who were unable to participate in some sessions still had access to project information.

Sessions followed a pattern of structured, task oriented, focused communication on detailed aspects of the project (e.g., “we should develop a front-end in Access that will allow the users to do simple things”). These sessions were used to assess tasks to be accomplished, coordinate project ideas, and establish responsibilities (e.g., “Sounds good in theory, let’s explore the different possibilities locally in the group and get back to you on choice of language”). These conversations were also used to establish timelines for the completion of tasks (e.g., “I’ll do both then, expect it around 24 hours.”) Chat communications were sometimes used to express opinions on previous suggestions; (e.g., “someone mentioned writing up a contract [with the client] that goes into more detail. I think that would be a good idea.”) Table 2b : Behaviors of Virtual Teams according to archetype (continued) Social Dimension High Low Time-Space Low SOCIAL BEING TEAMS These teams understand each other’s norms, values and experiences. They trust and rely on each other sharing humor and personal stories. Social Beings may not have well-defined common goals and they may not possess the necessary knowledge and skills necessary to complete a project.

These teams focus less on coordinating tasks, deliverables and timelines. They interact easily with each other’s time and space differences due to respect for others.

Progress to the mutual communication stage mostly to build social relationship and cohesion among team members. APATHETIC TEAMS Team members don’t fully understand each other’s norms, values and experiences. They lack trust, share little humor and few personal stories. Often these teams lack a common overall goal. They don’t possess the necessary domain knowledge and skills required.

Apathetic teams are not well coordinated in their tasks, deliverables, and timelines. They lack substantive interactions in both intensity and content and they have difficulty interacting. Team members are not sensitive to each other’s time and space differences. These teams never achieve mutual communication among team members, thus not being able to build social relationship or work effectively on the project. Table 2b. Behaviors of Virtual Teams According to Archetype (Continued) Systems Development by Virtual Project Teams 455 Copyright © 2003, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited. Copyright Idea Group Inc. Copyright Idea Group Inc.

Team B—Sport and Recreation Division Team B’s objective was to develop a Web-based scheduling system for the sport and recreation division. Initial communication between the teams occurred via a discussion posting. This initial communiqué was strictly task-related and offered no introduction to the project or team. Subsequent introductory messages by the team members followed a similar pattern of focusing on details of the client or project. Personal information about team members was not offered or shared by the members.

Messages posted to the discussion section of the collaborative tool were very short and to the point during this phase of the project, often only consisting of a few words (e.g., “I have posted proposal and client information in the shared files section of WebCT™ Post any questions you may have”; “I don’t see a file posted. Could you try again?”). Posts almost always focused on project related topics. Most posts didn’t even include salutatory introductions.

Team B utilized the chat capabilities of the web collaboration tool and chatted at least once a week. During the first phase of the project, chat sessions were used to clarify project details and establish rules of communication. The members initially used sessions to assign and clarify team roles. Chat sessions usually began with greetings, but quickly transitioned to conversations regarding project topics. Generally one of the team members suggested, “We should probably talk about the project now.” Or, “okay, we need to talk about the deliverable now.” Such transitions effectively cut off all social interactions. Conversation rarely reverted to social communications of any kind other than to say, “Good bye, have a good rest of the day.” Nine days after the first communication posting, a Norwegian member of the team posted a message containing personal information: “A few facts about myself: I am a girl - I do not think my name is used in USA, so this might be useful information. I am studying for a master’s degree. I will hopefully finish my master thesis in December—almost one year from now. I am writing a paper on estimating work in software projects by using UML Use Cases.” This message was followed by similar postings from two other members of the team.

Of the ten-member team, only four members provided any substantial personal information to the team throughout the duration of the project.

After one month, a team member from the United States posted the following message:

“I must unfortunately tell you all that I cannot continue with the virtual project. I have been assigned to other projects that require my full attention at this time. I hope that you understand. I wish you all the best of luck and success in the future.” Only one member of the team bothered to respond to the post saying simply, “Good luck to you.” Apart from this message and the personal information shared by a few members of Team B, all message posts focused on the project and contained no social communication.

As project deadlines approached the frequency of discussion postings increased.

Messages focused on which members would complete which parts of the next project deliverable. Questions regarding clarification of client expectations or information to be included in the deliverable were often asked via discussion postings. These sorts of “informational” posts were usually answered promptly. Team C—Envision, Inc. Initial interaction within the team occurred via messages posted to the discussion forum. Messages originated from Norwegian members who provided brief introduction of themselves, limited personal information, and contact information. Members from the United 456 Croasdell, Fox & Sarker Copyright © 2003, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited. Copyright Idea Group Inc.

Copyright Idea Group Inc.Copyright Idea Group Inc. States replied to the postings by providing similar information as well as expressing excitement for the project—“I just wanted to say we all look forward to working with you.” Discussion postings continued over the following weeks. For the most part postings were short, establishing rules and norms for the project. Most of the communications were uni- directional whereby one side initiated the communication with the other watching, feeling uncomfortable to interact, or simply indicating their own presence on a particular communi- cations channel (Lau et al., 2000).

After a few weeks, the team planned its first interactive chat session. The session focused on defining tasks necessary to successfully complete the project. Team members exchanged information regarding the project —“Well, we can create ASP web pages to access the information from a local database, or we could create a program that would allow them to enter info as well as extract it.” These sessions represented the first real use of bi-directional communication between team members. In bi-directional communication, both local and remote members talk past each other in exchanging task or socially related information.

Information is shared but responses are not meaningful and members’ circumstances and priorities are not considered (Lau et al., 2000). In fact, members ostensibly engaged in chat where not devoting much attention to the chat—“Sorry, I got distracted—now I am back” “Sorry about that I am at work;” “ahh, I am at work too.” Similarly, a few conversations occurred at the same time creating cross-talk and indicating a lack of attention/care to the ideas and questions being provided by different team-members.

Later conversations involved the sharing of personal information, humor and stories.

The group discussed upcoming members’ weddings, personal interests, even public transportation. Conversations such as these seemed to establish a social foundation and allow group member to move into a mutual communication pattern evident in the high degree of social cohesion in the group.

Team D—Nembus Following introductions, the team worked to establish a convenient time to hold synchronous chat sessions. Unfortunately, the group was unable to determine a time when all members of the team were available to participate. In the end, a time was selected that excluded at least two members from the online discussions. Prior to the first chat session, a member of the Norwegian contingent posted a message to the discussion forum. The post provided an agenda for the chat session. One of the agenda items indicated members of the U.S. contingent were to design and build a prototype for the system. A deadline for completion of the task was also provided. Team members from the United States had no prior knowledge of this prototype. One U.S.-based member stated, “I felt as if the Norwegian members were delegating project tasks to us without seeking information or collaboration from the United States-based members. They were assigning us tasks to complete without seeking any knowledge of our capabilities and/or skills. It was frustrating to be told to complete certain tasks. Had they bothered to ask, they would have discovered we did not have the required skills or knowledge to complete the task as assigned.” This chat was task-oriented with little social communication. Early in the chat, Norwe- gian members took a strong leadership role dictating how United States members would be participating in the project. The following comment from a Norwegian member is represen- tative of the entire dialogue: “One of the deliverables is a design document, how do you plan to develop this document?” “When making the design you will not specify a programming Systems Development by Virtual Project Teams 457 Copyright © 2003, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited. Copyright Idea Group Inc. Copyright Idea Group Inc.

language.” Frustration in the United States group developed because they felt as if they did not have a say in the project. According the U.S. contingent, ‘the Norwegians had defined the role of U.S. members as laborers whose primary job was to focus on the grunt work of the project.” By the second chat, frustrations had swelled. This feeling was magnified when United States members asked a question regarding an upcoming project deliverable. Norwegian members responded by saying, “I thought that was our deliverable.” This interaction provided the first indication that the two groups were destined to participate in the project as two separate teams working on one project. The attitude was prevalent for the remainder of the project and proved a serious obstacle in the team’s ability to successfully attain project goals. For the remainder of the chat, the members of United States asked questions regarding the project and expressed confusion regarding the answers provided. One member from United States commented, ‘We expressed the desire to learn more about the client and the project and were told the Norwegian members would take care of that area. This was extremely frustrating because we felt we didn’t have any say in the project and were just being delegated work.” Following this first chat, team members from the United States contacted project facilitators and expressed concerns that they were not being provided the opportunity to contribute to the overall project. They perceived that Norwegian members felt that United States members’ skills were inadequate to contribute in any significant way. When facilitators communicated this concern to the Norwegian team members, they quickly responded via a discussion posting. One member of the Norwegian team stated, “I am really sorry if you got the impression yesterday that your skills are not adequate.” A more in-depth explanation of the project followed but did not address issues of collaboration and teamwork. At best, the comments only added more details as to what Norwegian members expected of the United States members. Phase II: Design and Development The purpose of this phase in the development of the information systems projects was to analyze details of the current operations and outline problems, bottlenecks, opportunities and other issues present in the existing systems. Logical and physical design documents were created to layout system requirements and design specifications. Project teams develop strategies for acquiring or building the new systems described in Phase 1. Finally, project teams identified functional specifications such that programmers could develop the required systems. Project deliverables from this phase of the system development required significant investments in time and effort on the part of all team members. Team A—Aerial Painters, Inc. As the project moved into Phase II, discussion postings continued at a steady pace.

Posts were used to convey new information about the project and chat transcripts. However, after about six weeks, the communication began to change significantly. The team began to test the social aspects of the group to help understand the personalities of virtual members.

This shift from task communication to social communication was first evident during a chat with a conversation regarding “Americans and lawsuits.” Both sides joked about the potential for the system’s vulnerability to lawsuits and that it might be wise to hire Johnny Cochran to represent the group. This conversation was followed by emoticons to convey 458 Croasdell, Fox & Sarker Copyright © 2003, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited. Copyright Idea Group Inc.

Copyright Idea Group Inc.Copyright Idea Group Inc. personal emotions (e.g., LOL, ;-), hehehe). After a short discussion the conversation was refocused and the group “got down to business.” Similar instances of group members using humor to help uncover the personalities of members followed in the chats to come. The group joked about the time of day the chats were taking place, Member A - I’d say owl for us :); Member B - owl? Member A - It is pitch dark here; Member B - Bat time for you = owl time for us. (No big deal LOL). By the middle of the second phase of the project these exchanges of jokes and humor had become a significant part of chat sessions. Most chat session began with a few extensive conversations of humor and joking around before any actual project topics were discussed. This pattern continued throughout the remainder of the project. As the team members began to understand each other’s personalities the exchanges became more tailored to the various individuals. One member of the team commented that this “Shift to a more social communication pattern created an environment for a more open form of communication. The environment was then one where members could be constructively critical of ideas and thoughts without the fear of stepping on the toes of group members. This new social attitude also allowed for an environment in which ideas flowed with more freedom. While this didn’t add much efficiency to the chats, the content of the conversations was much richer.” Team B—Sport and Recreation Division By the second phase of the project, discussion postings had grown in length and detail.

Messages were typically specific to the project, frequently including detailed questions for specific members of the team. Roles and task assignments were assigned and posted as deliverable deadlines approached. Members of the team completed assigned tasks individu- ally and asked questions to group-members as necessary. Chat sessions continued to take place at least once each week. Unlike the chat sessions in Phase I, little or no small talk occurred. Conversations were strictly task-based.

Project progress and communication slowed down as United States members took time off for vacations. Upon their return, team momentum lagged. Frustration within the team grew becoming obvious when miscommunications regarding task assignments were uncovered:

“The screen designs was our task … it is unnecessary that both local teams spent time doing this. At the chat meeting we said you could send or fax what you’ve made so that we could get some ideas. You were not supposed to complete this task—this was a task assigned to the Norwegian team.” This message did not elicit a response from any member of the U.S.-based team. In fact, no action was taken to address the miscommunication and/or misunderstanding. Team C—Envision, Inc. As the project progressed through the first phase and into Phase II, Team C continued to utilize all of the collaboration tools available in WebCT™ Messages were frequently posted to the discussion forum. These messages usually consisted of short bits of information such as updates on deliverables status, system status or schedule of chat sessions. Substantial discussion rarely took place via the discussion forum.

Chat sessions remained the primary method of communication. However, after several weeks of working together, the chat sessions began to shift drastically from the task Systems Development by Virtual Project Teams 459 Copyright © 2003, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited. Copyright Idea Group Inc. Copyright Idea Group Inc.

orientation present in Phase I to an orientation that included much more social communica- tion. One session consisted entirely of conversations about the ages of team members, birthdays, the structure of higher education systems, recent and future travels, even wolves and cougars—“Cougars are dangerous, I think. Wolves are not, not to humans anyway.” These dialogues played an important role in the group’s ability to understand each other’s norms, values and experiences as well as shift the group communication pattern from bi- directional to mutual communication in which group members were “talking to each other” in a substantive fashion. The shift in orientation illustrated team members respect for each other and demonstrated the ability of each team member to consider individual circumstances later in the project.

At one point, social conversations encompassed approximately 75% of the chat session conversations. At times, non-project-related discussions dominated the dialogues so much that project tasks were not being addressed. In one instance a group member wrote, “I have to go in 15 minutes…could we go on?” After the request a few project points were discussed, but the conversation quickly reverted to more social topics. This pattern of putting social conversation first continued through phase two. Evidence of the lack of task focus began to frustrate some team members. In one session, members had to ask three times if they could begin discussing the project before the group began. “I think it is important that we decide it pretty quick as we don’t have much time left,” “ok…we should move on,” “we need to move on…” Misunderstanding and confusion also surfaced as social conversation obscured project tasks – “sorry about the confusion, it made sense to us, but I guess we didn’t think it through well enough.” Team D—Nembus By the second phase of the project team, cohesion had declined and frustration was increasing among members of Team D. Due to scheduling difficulties, some members of the team had yet to attend a single chat session. Consequently, they were unknown to remote group members. Progress was slow due to the lack of communication between group members. Six weeks into the project, the team used a chat session to finally set standards for document control. Shortly after this session, the problems that had developed within the team were brought to the surface and discussed. During one of the chat sessions a Norwegian team member expressed, “[United States]: we are sorry for not including you more in the earlier deliverables, was that a problem for you?” A U.S.-based team member responded, “Well, we’d like to start working on the prototype now.” It became evident that roles had not defined very well. Communication was lacking on both sides. Norwegian members felt they were commu- nicating while United States members felt they were on hold, waiting to be told when to participate. In fact, team members from the United States weren’t listed as team members on project deliverables, nor did they have assigned roles. Both sides of the group agreed to communicate more often and become more actively involved.

Late in the second phase of the project, it appeared the team had come together to work as one entity rather than two. The group had an extensive dialogue regarding the next deliverable and how the work was to be divided. Initially, Norwegian members suggested that United States unit do the majority of the work and they would help where needed, (“We have also been looking at the next deliverable. Hopefully, we can take some of the work off you.

However, this is only if you want to. We are more than confident that you can manage!”) A few discussion postings later this approach had changed and Norwegian members proposed 460 Croasdell, Fox & Sarker Copyright © 2003, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited. Copyright Idea Group Inc.

Copyright Idea Group Inc.Copyright Idea Group Inc. to work on six aspects of the deliverable and suggested United States produce the remaining two. All members of the group agreed that the matter should be discussed further and scheduled a chat session during which all members would attend. Unfortunately, all nine members of the group failed to recognize that time changed due to daylight savings time.

Groups arrived to chat at different times.

The miscommunication allowed each side of the group to chat among themselves as they waited for the remote members to arrive. During one of these conversations, a Norwegian member discussed the quality of output produced by their United States counterparts— “what they have done is repetition of what we have done earlier.” This comment led a few members of the group to becoming possessive. One Norwegian mentioned “our text” and asked, “why can’t they just refer to our old deliverable?” At this point in the conversation, a team member pointed out, “We are suppose to work as one group, not two.” It was clear to everyone that this principle was not being put into action by the team as a whole. Phase III: Implementation and Close-down The final phase of the project involved writing code according to specifications defined in Phase II and documenting the system for future users. Teams developed working systems with documented code, test procedures, test results and maintenance and users manuals. The virtual teams delivered final project deliverables via ISDN lines and Microsoft NetMeeting.

Videoconferencing allowed all team members to participate in showcasing their work. Upon delivery of the final project documents, team members were asked to reflect on their experience working in a virtual cross-cultural environment.

The third phase required teams needed to be highly interactive in order to accomplish all required tasks in the short time frame available to complete the projects. The stress was clearly evident in some groups, less so in others. Team A—Aerial Painters, Inc. The project team worked diligently to complete the system for the client. By the end of April, the group had successfully created a system that met and exceeded user-defined needs.

Over all, the group expressed excitement with the final results and that it was delivering a quality product. All members of the team participated in the final presentation. The videoconference was the first time group members interacted in a face-to-face environment.

However, team members had interacted frequently throughout the project using discussion forums and interactive chat sessions. Team members were familiar with personalities, norms, and values. Consequently, the group was able to deliver a presentation as one group working toward the same goal. The final project presentation was entirely task oriented. The usual jokes and pats on the back were not evident to the audience. Nonetheless, following the final presentation and delivery of the system to the client, all members posted messages to the discussion forum echoing congratulations on a great project and gratitude for the hard work done by all members. Team B—Sport and Recreation Department As the project entered the third and final phase, Team B’s communication remained largely task-oriented. Attempts at social interaction were made when Norwegian members of the team posted pictures of themselves working on the project. However, only one member of the U.S. contingent responded to the posting—seven days after they were initially posted! Systems Development by Virtual Project Teams 461 Copyright © 2003, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited. Copyright Idea Group Inc. Copyright Idea Group Inc.

Chat session occurred more frequently in this phase of the project as the team prepared for system delivery. Frustrations that had not been previously addressed continued to escalate over time leading to a near breakdown. Frustrated with the lack of communication and involvement coming from U.S. members, Norwegian members wrote a note expressing their frustrations. When members in the United States read these comments, one member posted a personal apology. “I want to start of by apologizing for my lack of communication over the past few weeks.” The team member went on by reassuring the group of her commitment to the project—“I am willing to put in as much time as needed to make this project successful so please let me know anything I can do and I will get it done.” The post drew a response from one member of the team and participation increased after the exchanges.

Team B presented the final product using NetMeeting (application sharing) and ISDN- based videoconferencing. After the presentation, team members posted messages to the discussion forum expressing their thoughts about the final project and the over-all experi- ence. “Nice to finally talk to you all ‘face-to-face’,” “I have had a great time working on this project and getting to know everyone. I think we did an amazing job and pulled off some great work.” Follow-up posts were scant and never drew responses from other team members.

Team C—Envision, Inc. As the group moved into the final phase of the project, they found themselves in a rush to finish the system as they had spent little time focused on task details in the second phase of the project. Three months into the project, they were finally able to declare that they “have a good understanding of the project now,” and “should be able to handle things from here.” The late start on project work caused members to become frustrated and anxious—“we are experiencing some set backs here … kind of frustrating for us.” The massive social interactions from previous phases seemed to be pulling the team apart. Fortunately, they played an important role in keeping the team together. In a weekly chat session, team members discovered that the client organization had gone bankrupt. Group members expressed disbelief and surprise, but the social relationships served to bond team members. Despite the clients impending demise, Team C made a decision to continue the project and deliver a solution reasoning that their solution could be useful as part of a future system. The individuals facilitating the virtual team environment agreed to continue with their sponsor- ship as well. One team member expressed her hope that the team could “manage to keep the spirit and complete the system even with the bad news.” Others replied that the team should be “motivated by the challenge of the system, not delivery of the final project.” All members agreed that they would make the best of a worse case. Here, the social bonds established in earlier project phases worked to hold the team together even the midst of a significant negative event.

Despite monumental eleventh-hour efforts to develop and code the applications, the team had produced a system that only partially met the specifications originally outlined by the client. Evidence of the social bond that had developed during early stages was evident in the final videoconference in which the both sides of the team interacted with comfort and ease. After the final delivery, members of the team posted closing remarks to the discussion forum expressing gratitude for the hard work and well wishes for the future. Team D—Nembus By the final phase of the project, the team had almost completely split into two separate teams. Distrust and dissatisfaction led the team to conclude their best alternative was to try 462 Croasdell, Fox & Sarker Copyright © 2003, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited. Copyright Idea Group Inc.

Copyright Idea Group Inc.Copyright Idea Group Inc. and deliver a prototype of a working system to the client. It was evident a full-fledged working system simple wasn’t in the cards. Ironically, Norwegian members had assigned the task of building a prototype to United States members early on in the project. In fact, one United States member dedicated substantial time and effort to developing the prototype using information from deliverables and information provided in chats and discussions. Unfortu- nately, scheduling conflicts and the lack of interaction in virtual space resulted in a prototype that was inconsistent with system requirements. After many long hours and significant struggle, the U.S. contingent was able to develop a prototype. Meanwhile, Norwegian members turned their attention to developing a final presentation for delivery of the project.

The group divided the presentation - each side focusing on topics they were familiar with.

Norwegian members discussed the analysis phase of the project. Members from the U.S.

focused their presentation on the prototype. As a result of their efforts, U.S. team members were able to present a “successful prototype.” However, when evaluated on communication and group cohesion evidence throughout the duration of the project, “success” appeared to be limited. In fact, the existence of any prototype at all was largely attributable to one or two individuals who took it upon themselves to develop a satisfactory product.

CHALLENGES OF VIRTUAL TEAMS IN SYSTEMS ANALYSIS AND DESIGN Virtual teams are quickly becoming an integral part of the business world adding functionality to group work as well as increasing competitive advantage. This study of virtual teams and the communication within helps enhance our knowledge and understanding of virtual teams providing insight into a participants’ views as well as providing information that would be useful in virtual team management. Information presented in the study can be used to help management build successful teams and can be used as a learning tool for future virtual teams aiding in their success. Determinants for success may lie in the shared frame of reference of the virtual teams, the ability and willingness to work through cultural differences, the capability to recognize differing skill and interests and the ability to incorporate technology to span time and space.

The cases presented in this comparative study demonstrate that successful utilization of communication tools helps develop strong social connections among team members. In part, these social connections allow team to deliver quality products to their clients. Frequent communication can create strong social bonds and allow team-members to grow together. For some virtual teams, the lack of social glue can create project breakdown. Teams that lack a strong presence of both social communication and task-related communication have more difficulty successfully completing projects. Lack of social interaction can create obstacles that are difficult to overcome in addition to creating difficulties in collaborating on the technical requirements for the project.

Questions still remain regarding the understanding of virtual teams. How do managers evaluate virtual teams? How do project leaders instill the factors of success in teams? What factors of virtual teams contribute to team productivity? As technology advances, how are virtual teams affected? Does “rich” communications actually lead to more productive cohesive teams? As virtual teams become increasingly important for businesses and are studied in greater detail, we will hopefully begin to gain some insights regarding some of the above issues. The review questions present additional areas for consideration and analysis. Systems Development by Virtual Project Teams 463 Copyright © 2003, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited. Copyright Idea Group Inc. Copyright Idea Group Inc.

REFERENCES Briggs, R.O., Nunamaker, J.H., & Sprague, R.H. (1998). 1001 unanswered research questions in GSS. Journal of Management Information Systems, 14(3), 3-21.

Gersick, C.J.G. (1988). Time and transition in work teams: Toward a new model of group development.Academy of Management Journal, 31, 9-41.

Jablin, F. M. & Sias, P. M. (2000). Communication competence. In F. M. Jablin & L. L. Putnam (Eds.),The New Handbook of Organizational Communication: Advances in Theory, Research, and Methods, 47-77.

Lau, F., Sarker, S., & Sahay, S. (2000). On managing virtual teams. Healthcare Information Management & Communications, 14(2), 46-52.

Sahay, S. & Krishna, S. (2000). Understanding global software outsourcing arrangements:

A dialectical perspective. Working Paper, Indian Institute of Management, Bangalore, India.

BIOGRAPHICAL SKETCHES David Croasdell is an assistant professor of Management Information Systems in the School of Accounting, Information Systems and Business Law at Washington State University, Pullman, USA. His current research interests include Organizational Memory, Knowledge Management, and Inquiring Organizations. Dave teaches courses on Systems Analysis and Design, Data Communications and Networking, and Electronic Commerce.

Andrea Fox recently graduated from Washington State University, USA, with a degree Business Administration. Her major area of study was Management Information Systems. Andrea’s honors curriculum involved intensive participation and study of cross- cultural system development virtual teams. Andrea currently works in technology risk consulting with a global professional services firm.

Suprateek Sarker is an assistant professor of Management Information Systems in the School of Accounting, Information Systems and Business Law at Washington State University, Pullman, USA. His current research interests include virtual teamwork, IT- enabled change, and on-line education.