The decoy effect describes a bias in which people's choices between two valuable options are swayed by a third, inferior, "decoy" option. Despite being documented in lab settings, relatively little work has investigated whether decoy effects occur "in the wild" where consumers face large, diverse choice sets. We employ a new methodology to examine the impact of decoy options on purchase decisions using a dataset of 3.6 million UK grocery-store wine transactions. Results indicate that when comparing wines that vary in quality and price across contexts, the presence of dominated (i.e., inferior) decoy options increased consumers' likelihood of choosing a target option--a hallmark of the well-documented attraction effect. The strength of these effects was modest overall (roughly 1% change in preference) and, interestingly, depended on consumers' idiosyncratic histories of experience. Our study provides a proof of principle demonstrating that these sorts of context effects are detectable in richer, complex real-world consumer choice settings.