We use a multi-method approach (analytical model and behavioral experiment) to investigate product recommendations based on less-important attributes (weak unique selling proposition, USP). We consider multiple scenarios in which a recommender’s level of expertise (knowledge about product attributes and their importance) and bias (preference for the firm as opposed to consumers) operate as cues for consumers to evaluate the recommender’s message. Results show that optimal messaging behavior is a function of an interactive process involving recommender characteristics and the relative importance of product attributes to consumers. The results identify conditions that determine when weak USPs are likely to increase or decrease a consumer’s propensity to buy the recommended product and when a recommender might optimally communicate weak USPs or avoid sending such a recommendation.