Algorithm‐Generated Identity Labeling Promotes Identity‐Consistent Product Preferences

• Published in: Psychology & marketing Vol. 42; no. 6; pp. 1667 - 1679
• Main Authors: Shen, Zhengyu, Hu, Yunxuan, Yin, Yunlu
• Format: Journal Article
• Language: English
• Published: Hoboken Wiley Periodicals Inc 01.06.2025
• Subjects: Algorithm; Algorithms; Consumer behavior; Consumers; expected accuracy of algorithmic prediction; identity label; strength of identification
• ISSN: 0742-6046; 1520-6793
• DOI: 10.1002/mar.22195

ABSTRACT A surging amount of companies have employed algorithms to generate identity labels for consumers, aiming to attract their interest and boost their engagement in marketing activities. However, it remains unclear whether consumers' social and product preferences may be shaped by the identity labels generated by the platform, depending on the source of these labels (i.e., whether identity labels are generated by algorithms or humans). Three studies covering different scenarios involving various identity labels and their corresponding identity‐consistent products show that algorithm‐generated (vs. human‐generated) identity labels enhance consumers' preferences for identity‐consistent products. We further demonstrate that this effect occurs because consumers exposed to algorithm‐generated identity labels exhibit a heightened expected accuracy of algorithmic predictions and a stronger strength of identification, which subsequently fosters their identity‐consistent product preferences. Our findings advance the understanding of consumer responses to algorithm‐generated information and offer valuable insights for companies to leverage algorithmic technology to nudge consumer behavior.