Searching in Teams: An Experimental Study of Aggregation and Incentives

09. Dezember 2013


Search is a ubiquitous activity in organizations. For example, organizations search for new employees, for process improvements, and for new products and services. Many of these search activities are the collective effort of teams. For instance, in product development, engineers with different technological backgrounds work on different components of the entire product architecture to improve performance. The broad contours of how teams collaborate in a search task are well-understood. In a team, the overall task of searching for improvements is divided and allocated to specialists, each focusing on the assigned subtask. The subtasks contribute jointly to overall performance and often interact non-trivially when generating the overall performance outcomes. This is because most tasks are imperfectly decomposable and subtasks exhibit strong interdependencies that can have significant performance implications. This problem is magnified by uncertainty about the exact nature of interdependencies. Prior work has highlighted the importance of thick communication, shared identity, and the coordinative capacity provided by heavy-weight team managers as vital instruments to facilitate team search. However, most of the prior empirical work focuses on outcomes, and there is a lack of knowledge concerning the process of team search. We address this gap with a laboratory experiment in which we study how teams engage in recombinant search, how their behavior and performance is shaped by incentives, and how team search is different from search that is conducted by individuals.