The impact of humans vs. AI recommendation on consumer reactions to products exposure

Abstract This study compares consumer reactions to product recommendations provided by AI versus expert human agents for search and experience products. Across three experimental studies, we propose that the effect of recommendation source and product type on intention to follow the recommendation is explained by recommendation source’s perceived transparency and credibility. We demonstrate that AI is perceived as more transparent and credible than a human expert when recommending search products, leading to a higher intention to follow the recommendation, while no difference emerges for experience products. However, when the human recommender is described as a Super Expert – highly experienced, reputable, and qualified – consumers show a preference for the human (vs. AI) source in the case of experience products, while the difference between the two recommendation sources became nonsignificant for search products. Furthermore, when recommendations come from a hybrid source combining a Super Expert and AI, this combination is evaluated less favorably than either source alone for search products, with no significant difference found for experience products. These results offer valuable insights for marketers on how to select, design and deploy effective touchpoints that enhance the willingness to follow recommendations, depending on the product type.