Can Automated Group Recommender Systems Help Consumers Make Better Choices?

Abstract

Because hedonic products consist predominantly of experience attributes, often with many available alternatives, choosing the "right" one is a demanding task for consumers. Decision making becomes even more difficult when a group, instead of an individual consumer, will consume the product, as is regularly the case for hedonic offerings such as movies, opera performances, and wine. Noting the prevalence of automated recommender systems as decision aids, the authors investigate the power of group recommender systems that consider the preferences of all group members. The authors develop a conceptual framework of the effects of group recommenders and empirically examine these effects through two choice experiments. They find that automated group recommenders offer more valuable information than single recommenders when the choice agent must consume the recommended alternative. However, when agents choose freely among alternatives, the group's social relationship quality determines whether group recommenders actually create higher group value. Finally, group recommenders outperform decision making without automated recommendations if the agent's intention to use the systems is high. A decision tree model of recommender usage offers guidance to hedonic product managers.89 † Significant (p < .05) for two standard deviations. † † Significant (p < .05) for one and two standard deviations. Notes: n.s. = not significant. SRQ = social relationship quality, IN = intention toward recommender usage. The t-values are equal for slopes at group recommendation system and single/no recommendation system because these are dichotomous and thus SD-independent variables.

“Share and scare”: Solving the communication dilemma of early adopters with a high need for uniqueness

“…As presented in Appendix B (Supplementary material), Table B1 (first column), there was a significant two‐way interaction between early adoption and need for uniqueness ( β = .33, t (79) = 3.49, p < .01). To gain a deeper understanding of this interaction, we ran a spotlight analysis (Aiken & West, 1991; Fitzsimons, 2008; Hennig‐Thurau, Merchand, & Marx, 2012; Spiller, Fitzsimons, Lynch, & McClelland, 2013). Consistent with H1a, among early adopters (one SD above the mean), the dilemma was more intense when need for uniqueness was high (one SD above the mean) versus low (one SD below the mean) ( M = 2.69 vs. M = .77, β = .74, t (79) = 5.93, p < .001).…”

Recommender System Rethink: Implications for an Electronic Marketplace with Competing Manufacturers

“…Recommender systems have become the hallmark of modern eCommerce platforms. These systems are touted as sales support tools that help consumers identify their ideal product from among the vast variety sold by an eCommerce platform (Hennig-Thurau et al, 2012). It has been reported that over 35% of sales on Amazon.com and more than 60% of the rentals on Netix result from recommendations (Fleder et al, 2010).…”

Marketing Intelligence and Big Data: Digital Marketing Techniques on their Way to Becoming Social Engineering Techniques in Marketing

“…They generate personalized predictions about product liking, by filtering the past behavior and preference statements of consumers. [21] The social engineering technique, where trust is exchanged for reputation, became a digital version known as digital recommendation marketing, so that digital techniques and social measures are able to become two sides of the same marketing coin.…”