On matters of love: exploring consumer reactions to algorithmic vs. human dating recommendation agents

Abstract

Purpose Drawing on agency theory and implicit theory, this research examines how algorithmic versus matchmaker-provided dating recommendation service influence consumers' consideration of using the recommendation service. Design/methodology/approach This research conducted one pilot study (N = 233) and two scenario-based experimental studies (Study 1: N = 300; Study 2: N = 520) to investigate the relationship between agent type and service consideration. Findings The results show that agent type (algorithm vs. matchmaker) influences consumers' sense of control, which in turn determines their receptivity to dating recommendation service. Moreover, a situationally activated implicit belief in human intelligence (fixed vs. malleable) moderates this effect, such that stronger fixed beliefs are associated with greater consideration of services provided by algorithmic agents. Research limitations/implications This work contributes to understanding of consumers' appreciation for algorithms in online dating service by offering explanations for their inherent demands for agency, specifically the sense of control in individuals' experience during dating recommendation service. Practically, it provides valuable guidance for tailoring service delivery to users' need for control and agency. Originality/value This research documents a reversal of the well-established algorithm aversion in subjective tasks. Specifically, when algorithmic and human agents are framed as equally effective, consumers in the online dating context exhibit greater willingness to use algorithmic (vs. human) recommendation services.