Optimal Monitoring in Salesforce Control Systems

In an earlier contribution on salesforce control (Joseph and Thevaranjan 1998), we examine the impact of introducing monitoring on the optimal level of risk-aversion desired in the salesperson and the optimal commission rate offered to him (or her). There, however, we do not solve for the optimal level of monitoring because of the complexities associated with defining the level of monitoring and the difficulty of introducing the cost of monitoring into the analysis. Thus, in this paper, we extend our earlier analysis to an N-activity setting and complete our line of inquiry by explicitly solving for the optimal level of monitoring. We find that the optimal level of monitoring decreases as the output measure becomes more informative about salesperson effort but increases as both the efficiency of the monitoring technology possessed by the firm and the marginal cost of risk-tolerance in the labor market increases. In addition, a key benefit of our N-activity model is that it allows for a more precise interpretation of the results documented in Joseph and Thevaranjan (1998). For example, readers examining Joseph and Thevaranjan (1998) may erroneously conclude that increased monitoring will always lead to a lower weight on incentive pay. The "full" analysis presented in this paper, on the other hand, reveals that changes in the control system may sometimes involve both higher monitoring as well as a greater reliance on incentive pay.