Fast Success and Slow Failure: The Process Speed of Dispersed Research Teams

Marie Louise Mors*, David M. Waguespack

*Corresponding author for this work

Research output: Contribution to journalJournal articleResearchpeer-review

Abstract

Research teams are often dispersed across geography and organizational boundaries. While prior work has recognized that dispersed teams face coordination challenges, it is not clear how dispersion affects team process efficiency outcomes, such as the speed with which team efforts come to a resolution. Process efficiency outcomes are important because, if the performance benefits associated with working in dispersed teams have a trade-off cost in terms of efficiency, then the net performance benefits from working in such teams may be seriously diminished or even reversed. Prior work has also tended to focus on successful outcomes, which may lead to deficits in our understanding of coordination challenges in dispersed teams. To better comprehend process efficiency, we examine 5,250 teams that work together in an open standard setting, and where the time to resolution of both successful and failed projects is observable. We find that teams that are organizationally or geographically dispersed are fast at reaching success, but fail more slowly than non-dispersed teams. Our interpretation of these disparate outcomes is that the endogenous processes that make dispersed teams more likely to on average select higher potential projects have a second order effect of making it harder for dispersed teams to abandon failing projects. We argue that slow failure has important implications for research & development efforts given the opportunity cost of tying up resources in research teams.
Original languageEnglish
Article number104222
JournalResearch Policy
Volume50
Issue number5
Number of pages16
ISSN0048-7333
DOIs
Publication statusPublished - Jun 2021

Keywords

  • Research teams
  • Dispersed teams
  • Collaboration
  • Process efficiency
  • Coordination
  • Open source communities

Cite this