Phantoms are dominating, attractive alternatives that are unavailable at the time of choice. They occupy different positions in the attribute space, and their unavailability can be either known or unknown to individuals. Although different theories have offered explanations for the influence that phantoms exert on individual choices, they have largely overlooked phantom knowledge. Contradictory predictions can be drawn from these theories, and none provides a good account of the pattern of effects that emerges from our data. We suggest that these contradictions appear because the different theories do not address phantom location and knowledge jointly. When considering phantom knowledge together with location, we observe a consistent overall pattern of effects that encompasses all predictions that could be made on the basis of each theory. We find that known phantoms are stronger when close, and favor the target product, whereas unknown phantoms are stronger when far, and favor the competitor. Finally, we look beyond shifts in choice shares and find that phantom decoys can also affect individuals' post‐choice evaluations and reactions in terms of perceived justice, decision satisfaction, and repatronage intention. Our results show that post‐choice evaluations are driven by phantom knowledge, not phantom location. Copyright © 2012 John Wiley & Sons, Ltd.