This multilevel study examines the role of supervisors in improving employee performance through the use of coaching and group management practices. It examines the individual and synergistic effects of these management practices. The research subjects are call center agents in highly standardized jobs, and the organizational context is one in which calls, or task assignments, are randomly distributed via automated technology, providing a quasi-experimental approach in a real-world context.

more »

... lts show that the amount of coaching that an employee received each month predicted objective performance improvements over time. Moreover, workers exhibited higher performance where their supervisor emphasized group assignments and group incentives and where technology was more automated. Finally, the positive relationship between coaching and performance was stronger where supervisors made greater use of group incentives, where technology was less automated, and where technological changes were less frequent. Implications and potential limitations of the present study are discussed. This multilevel study examines the role of supervisors in improving employee performance through the use of coaching and group management practices. It examines the individual and synergistic effects of these management practices. The research subjects are call center agents in highly standardized jobs, and the organizational context is one in which calls, or task assignments, are randomly distributed via automated technology, providing a quasi-experimental approach in a real-world context. Results show that the amount of coaching that an employee received each month predicted objective performance improvements over time. Moreover, workers exhibited higher performance where their supervisor emphasized group assignments and group incentives and where technology was more automated. Finally, the positive relationship between coaching and performance was stronger where supervisors made greater use of group incentives, where technology was less automated, and where technological changes were less frequent. Implications and potential limitations of the present study are discussed.