تیم های موازی برای ایجاد دانش : نقش همکاری و مشوق ها

Parallel team strategy, in which multiple teams simultaneously pursue project goals, has been widely adopted by high-tech industries for knowledge creation. In this study, we investigate the design of organizational incentives, including a fixed wage payment and an additional reward structure, for effective management of the parallel team strategy. We consider two main variants of parallel teams—collaborative and non-collaborative teams. Proposing and investigating three types of organizational reward policies, individual, aggregate, and contingent, we analyze the viability and characteristics of these policies. We show that individual reward policy performs better than aggregate policy, and that collaboration in parallel teams is vital. When parallel teams work non-collaboratively or when aggregate reward policy is used for collaborative teams, the firm achieves optimal profits by only offering a share of the knowledge creation benefit as the reward. Under some conditions for collaborative teams, we demonstrate that individual and contingent reward policies can achieve maximal benefits (first-best) for the organization. This research provides valuable insights for firms in employing parallel team strategy for knowledge creation.

Teams serve as effective organizational structures for pursuing innovation in firms. In a ground-breaking study of the practices of General Motors in 1943, Peter Drucker pointed out the effectiveness of team-based structures in organizations. As evidenced by numerous successful cases, team structures have been employed in many new projects. For example, Microsoft launched its gaming platform Xbox within a short period of time by employing the team strategy. At Google, various teams are assembled to work on different projects, such as Google Documents, Google Health, and Google Checkout. Past research provides insightful perspectives on team composition. For instance, Drucker [9] identifies three kinds of teams as baseball, football, and tennis doubles teams. Cohen and Bailey [6] categorize teams in organizations into four types: work, parallel, project, and management teams. Katzenback and Smith [20] suggest three types of teams: teams that—recommend, do, or run things. These papers share some similar notions on teams, for instance, the parallel team categorization of Cohen and Bailey [6] is the same as the football team identified by Drucker [9]. Taking into account the increasingly complex nature of technologies, recent IS research on teams has considered virtual teams, for instance, the application of group support systems in supporting them [18] and the social factors for their creation and effectiveness [23]. Not only do firms have to adopt appropriate team structures, but also assemble multiple teams to collectively engage in a single project. For example, Microsoft organized more than 200 programmers into several teams in developing the Windows95 operating system which contains more than 11 million lines of code [8]. Firms are continuously seeking cost-effective strategies to manage and coordinate teams on innovative projects. Different team strategies have been employed by industries for innovation and research. Concurrent team strategy is widely adopted to shorten the development time of new products. For instance, project managers at Microsoft usually divide a project into modules and different teams then work simultaneously on their modules but they synchronize with each other and debug daily [7]. In contrast, in the parallel team strategy, many teams work on the same research project simultaneously so as to maximize the overall success rate of the project. We cite a broad range of applications where parallel team strategy has been successful. It is frequently applied in research where single-team strategy results in high failure rates. For example, Nelson [29] documents the adoption of parallel strategy in R&D by the U.S. Air force. A similar parallel team approach was also used at National Institute of Health (NIH) to develop a malaria vaccine, instead of the traditional process of malaria vaccine development, where if one approach fails, the entire effort dies with it as well [32]. Arditti and Levy [2] report the use of parallel path strategy in the development of aircraft engines and MIG fighters, and also in the electronics industry for color TV development. Siegal and Chang [34] describe Samsung's use of parallel product development teams—teams in California and Korea working simultaneously pursuing different approaches in DRAM chip development, but cooperating as well. Wiklund [38] advocates the use of parallel drug development in which several candidate compounds are evaluated simultaneously to bring in productivity gains and shorter development times. In this paper, we model how to combine the parallel team strategy with incentives for effective knowledge creation. We highlight this paper's contribution by considering two related important dimensions of team functioning: (i) whether the teams work in parallel or not, and (ii) whether the teams collaborate or not. Most prior research has addressed primarily non-parallel teams focusing on issues such as team structure, team motivation, and social and organizational factors impacting team performance [19]. Some papers incorporating collaboration within and between non-parallel teams investigate how collaboration improves team performance [1]. Among the very few papers that have analyzed the parallel team strategy, Arditti and Levy [2] study how to choose the best number of parallel teams in new product development; but their context is limited to non-collaborative teams. In contrast, our paper primarily focuses on the so-far-unaddressed parallel and collaborative teams and highlights the role of collaboration by comparing it to the case when it is absent. Considering the parallel teams strategy, we study how the design of incentives and collaboration affect the success rate of innovation. Specifically, we address the following research questions in this paper. First, how does the design of teams differ between collaborative and non-collaborative parallel teams? Multiple parallel teams can be formed as either non-collaborative team or collaborative team similar to the differentiation between working groups and teams by Katzenback and Smith [19]: non-collaborative teams are loosely bound together for some common goals, whereas collaborative teams coalesce because of the collaboration among them. In particular, non-collaborative teams work independently without learning from or sharing with other teams, whereas collaborative teams work closely together so as to effectively increase the success rate of the research project. We study whether and how the presence of collaboration in parallel teams helps a firm improve knowledge creation. Second, how should a firm design incentive structures for parallel research teams? The incentive structure for a team in this research consists of a fixed wage payment and an additional reward policy. All the teams get the fixed payment irrespective of their individual success and they will be rewarded additionally if their research project succeeds. Three types of reward policies (individual, aggregate, and contingent) are proposed in this paper and we analyze how these rewards should be designed so that the firm may achieve maximal profit. Finally, how should the firm match the reward policy appropriately for different team structures? We examine whether different reward policies can achieve the maximal profit (first-best) for the two team structures. We show the conditions under which these reward structures are effective. Our paper makes several significant contributions. While it is true that collaboration can be expected to improve team performance, we demonstrate that the first-best solution can be achieved even when workers voluntarily choose their effort levels. Similarly, although rewards can generally improve performance, we characterize the structure of the reward itself. Moreover, we explore the characteristics of various reward policies—individual, aggregate, and contingent policies, including the appropriate number of teams. We show that individual reward policy performs better than aggregate policy, and that collaboration in parallel teams is vital. When parallel teams work non-collaboratively or when aggregate reward policy is used for collaborative teams, the firm achieves optimal profits by only offering a share of the knowledge creation benefit as the reward. In spite of the negative complementarity between the effort level and the total number of teams, we demonstrate that individual and contingent reward policies can achieve maximal benefits (first-best) for the organization. Therefore, managers can employ our results to design reward structures to improve knowledge creation outcome. The paper proceeds as follows. Section 2 briefly reviews prior related research, Section 3 outlines our model, Section 4 presents the detailed analysis and discussion, and Section 5 provides managerial insights and concludes the paper.

The increasingly competitive market has forced companies to seek more cost-effective ways to engage in knowledge creation. Recent trends in knowledge creation clearly indicate that companies are constantly searching for a better strategy to not only lower costs but also improve the quality of knowledge discovery. In this paper, we presented a model of parallel teams in which a firm employs multiple teams and designs incentives to motivate these teams to exert their best effort. We derive important managerial insights about the parallel team strategy and our analysis provides valuable guidelines for managers in deploying parallel teams as discussed below. First, incentivizing teams to effectively engage in knowledge creation is essential for firms to improve their productivity and overall performance. We show how to design incentives by choosing the combination of a fixed wage payment and rewards to induce workers' best efforts in knowledge innovation, enhancing the success rate of research projects for potential knowledge discovery. Second, when parallel teams work non-collaboratively due to organizational constraints, the firm should not offer any fixed payment to the teams, but only offer a share of the knowledge creation benefit as the reward. Third, under a collaborative structure, we demonstrate that the first-best solution can be achieved under certain conditions. Successful teams can be rewarded individually or all teams collectively when the project succeeds. Under the individual reward policy, although it is possible to achieve the first-best effort levels, the incentive to motivate collaboration may not be so strong because only successful teams get the reward. In contrast, under the aggregate reward policy, teams will equally share the total reward as long as a team succeeds, so the teams will collaborate voluntarily. However, the first-best solution cannot be achieved. To take the advantages of the positive features of both the individual and aggregate reward policies, we propose the contingent reward policy that rewards only successful teams and the amount of reward increases with the total number of successful teams. This contingent policy can be implemented to achieve the first-best solution and also facilitate the collaboration among teams. This study sheds light on how incentives and collaboration among teams affect organizational decisions on employing parallel teams in knowledge creation. Future research may explore the competition among parallel teams and how to exploit it for organizational benefit, study the uncertainty of innovation benefit with potential knowledge discovery, and investigate the impacts of information technology on team collaboration. In conclusion, our paper provides valuable insights for managers to determine the structure and level of rewards in managing parallel teams and achieving optimal organizational benefits.