Algorithmic Supervisor and Gender Difference in Worker Performance

Artificial intelligence (AI) has revolutionized the workplace. In this paper, we explore how an AI supervisor— a data analytics tool that provides real-time feedback to employees—affects employee performance and performance differences between male and female employees. We collaborated with a leading service company that began using AI to manage call agents in its outbound call center on September 18, 2019. We collected data for 220,085 calls over a span of three months before and after the introduction of the AI supervisor. We find that the introduction of the AI supervisor improved the agents’ service quality in terms of customer satisfaction, but this came at the expense of quantity as fewer customers were served. Moreover, we find that the adoption of the AI supervisor reduced agents’ performance inequality between female and male agents. In particular, before the introduction of the AI supervisor, male agents outperformed female agents, while after the adoption, there was an increase of 1.00 customers served by female agents per hour and an increase of 2.14% in female agents’ customer rating relative to male agents. Further, we find that this reduction in the gender performance gap is more pronounced for less experienced agents. Our research implies that the adoption of AI supervisors has an economic value on improving customer satisfaction and a social value on reducing gender differences between employees