No abstract

doi 10.1287/orsc.1070.029

versity. His research focuses on understanding how teams coordinate across geographic and global boundaries and which team processes and information technologies are most effective in bridging these boundaries to achieve high levels of performance. His current research areas include global software and technical teams, team knowledge, team coordination, and spatial and temporal boundaries. SANDRA A. SLAUGHTER is an Associate Professor in the Tepper School of Business at Carnegie Mellon University. Prior to joining the faculty at Carnegie Mellon University, Dr. Slaughter worked in industry as a project manager and systems analyst at Hewlett-Packard, Rockwell International, and Square D Corporation. She has consulted with the Information Technology Management Association and with several companies on software design and development-related issues. Her research is motivated by her experience in software development and focuses on the performance implications of software design and development decisions. Currently, she is conducting research funded by the National Science Foundation on project management practices and software design and evolution. She has also commenced new projects that explore capabilities and performance in information technology outsourcing.ABSTRACT: Coordination is important in software development because it leads to benefi ts such as cost savings, shorter development cycles, and better-integrated products. Team cognition research suggests that members coordinate through team knowledge, but this perspective has only been investigated in real-time collocated tasks and we know little about which types of team knowledge best help coordination in the most geographically distributed software work. In this fi eld study, we investigate the coordination needs of software teams, how team knowledge affects coordination, and how this effect is infl uenced by geographic dispersion. Our fi ndings show that software teams have three distinct types of coordination needs-technical, temporal, and process-and that these needs vary with the members' role; geographic distance has a negative effect on coordination, but is mitigated by shared knowledge of the team and presence awareness; and shared task knowledge is more important for coordination among collocated members. We articulate propositions for future research in this area based on our analysis. KEY WORDS AND PHRASES: coordination, global software development, management of the information technology (IT) function, team knowledge.LARGE-SCALE SOFTWARE DEVELOPMENT REQUIRES a substantial amount of coordination because software work is carried out simultaneously by many individuals and teams, and then integrated into a single product. Software parts need to integrate and interoperate properly, and production schedules need to be synchronized, creating dependencies among tasks and people. These coordination challenges are compounded when the teams doing the work are distributed across multiple geographic locations. Developing software globally is increasingly b...

Team s are important units of organizational work because they bring diverse expertise, skills, and resources to complex tasks that may be too large or complex for a single individual to undertake. However, as projects and teams grow in size and complexity, tasks and member dependencies become more numerous and more complex, thus increasing the need for team coordination. Effective teams manage these dependencies using a number of explicit and implicit coordination mechanisms and processes. Teams coordinate explicitly using task programming mechanisms (e.g., schedules, plans, procedures) or by communicating (e.g., orally, in writing, formally, informally, interpersonally, in groups). We call these mechanisms explicit because team members use them purposely to coordinate. However, teams can also coordinate implicitly (i.e., without consciously trying to coordinate) through team cognition, or knowledge that team members share about the task and

In globally distributed projects, members have to deal with spatial boundaries (different cities) and temporal boundaries (different work hours) because other members are often in cities within and across time zones. For pairs of members with spatial boundaries and no temporal boundaries (those in different cities with overlapping work hours), synchronous communication technologies such as the telephone, instant messaging (IM), and Web conferencing provide a means for real-time interaction. However, for pairs of members with spatial and temporal boundaries (those in different cities with nonoverlapping work hours), asynchronous communication technologies, such as e-mail, provide a way to interact intermittently. Using survey data from 675 project members (representing 5,674 pairs of members) across 108 projects in a multinational semiconductor firm, we develop and empirically test a relational model of coordination delay. In our model, the likelihood of delay for pairs of members is a function of the spatial and temporal boundaries that separate them, as well as the communication technologies they use to coordinate their work. As expected, greater use of synchronous web conferencing reduces coordination delay for pairs of members in different cities with overlapping work hours relative to pairs of members with nonoverlapping work hours. Unexpectedly, greater use of asynchronous e-mail does not reduce coordination delay for pairs of members in different cities with nonoverlapping work hours, but rather reduces coordination delay for those with overlapping work hours. We discuss the implications of our findings that temporal boundaries are more difficult to cross with communication technologies than spatial boundaries.