Technically feasible, ethically questionable? Psychological pitfalls of digital eavesdropping

Abstract

Digital eavesdropping refers to a little-known yet legally permissible data provenance technology that enables mobile devices to listen to personal (“face-to-face”) conversations, to identify keywords, and to deliver personalized advertisements to unsuspecting users. Unlike (re-)targeting, which relies on online browsing behavioral data provenance, digital eavesdropping generates personalized advertising based on spoken (offline) words. Across seven experiments involving various products and services, it is demonstrated that digital eavesdropping elicits heightened perceptions of creepiness – exceeding that of retargeting. This creepiness triggers adverse reactions to advertisements, including negative attitudes and reduced click-through, indifferent to trait variables. These effects are especially pronounced when ads are perceived as accurately tailored to consumers’ former voiced desires and, surprisingly, when consumers are explicitly informed about the ad algorithms and how digital eavesdropping operates. The findings suggest that the implementation of digital eavesdropping harms marketing effectiveness, should be approached with caution, and employed only under highly restricted circumstances.