Presenter: Qiaowen Guo, Washington University in St. Louis (link to personal website)

Field: Business Intelligence

Paper: Balancing Efficiency and Consumer Experience: An Investigation of Gig vs. Traditional Customer Support, https://dx.doi.org/10.2139/ssrn.5380791

Abstract: The rise of gig economy has transformed customer support, yet rigorous evidence comparing algorithm-controlled gig workers to traditionally-supervised employees remains scarce. We conducted a large-scale field experiment at a leading e-commerce platform, randomly routing consumer inquiries to either gig or traditional agents. While gig agents excel on algorithm-monitored metrics-achieving higher closure rates and faster service-they generate more negative consumer emotions and reduce future purchases. We identify strategic case transfers as the key mechanism: gig agents transfer cases at a much higher rate, with around half of them being unnecessary escalations used to boost performance metrics. This gaming behavior simultaneously inflates efficiency metrics and degrades consumer experience, ultimately eroding trust and leading consumers to shift future purchases from high-value durable goods to low-cost consumables. Our findings reveal how algorithmic control's narrow metric focus can backfire, suggesting platforms need more nuanced performance systems that balance efficiency with unmeasured quality dimensions.

Host: Allan Würtz