Do Activity-Based Incentive Plans Work? Evidence from a Large-Scale Field Intervention

Abstract

Many firms incorporate activity-based incentive (ABI) compensation into their pay plans. These ABIs are based on salespeople’s activity measures derived from their call reports. Despite their prevalence and theory-based expectations, there is a distinct lack of empirical work studying the sales productivity effects of ABI pay. With the cooperation of a large pharmaceutical firm, the authors conducted a three-year-long intervention based on a “treatment-removal” design. Their first intervention added modest ABI pay for frontline salespeople and their supervisors across 305 sales territories; the second intervention removed ABI pay from the salespeople; and the third intervention removed ABI pay from the supervisors as well, returning to the status quo. Using detailed territory-level data from the intervention in conjunction with syndicated market-level data and employing synthetic control procedures, the authors find sales gains of around 6%–9% from each of the two ABI interventions relative to the no-ABI baseline. These effects are moderated by the number of salespeople in a territory, with territories with more salespeople showing larger effects. Analyses of activity effects show that when supervisors are paid ABIs, they exert behavior control downward on salespeople. Managerially, both ABI schemes improve performance over an output-only pay plan. Between the two, a rudimentary gross profit impact calculation shows that ABIs targeted at supervisors alone are more efficient than ABIs targeted at both salespeople and their supervisors. The results support tying compensation to call reports despite the potential for self-serving biases in these measures because supervisors are able to exercise more behavior control with ABIs.