The effects of two coaching styles, one guidance and one facilitation, on the performance of coached and transfer tasks were examined in an experimental study. With the aim of improving and enhancing individual performance, guidance coaching entails the coach as a role model, delivering clear expectations and feedback about how to improve in a directive manner, while facilitation coaching involves the coach's helping the individual to explore and evaluate the task and self‐developing the correct responses for improving performance. Tests of the mediating effects of self‐set goals, which are defined as desired levels of performance to be attained on a task, and subjective task complexity, which is defined as an individual's perception of the complexity of a task, on the coaching style–performance relationship were also conducted. Participants (n = 127) were coached in the use of two software programs, PowerPoint and Excel 2007. The results showed that guidance coaching is more effective for improving coached task performance than facilitation coaching, while facilitation coaching is more effective for improving transfer task performance than guidance coaching. Subjective task complexity and self‐set performance goals mediated the effects of coaching style on the performance of both coached and transfer tasks. Implications for theory and managerial practice are discussed.