Past research (Kruger, Wirtz, Van Boven, & Altermatt, 2004) proposed that people use the effort of the producer as a heuristic for the quality of the product. In contrast, two experiments show that consumers' inferences from effort information are highly malleable. Participants were either explicitly exposed to one of two applicable naive theories ("good-art-takes-effort" vs. "good-art-takes-talent") or the order of judgment was reversed (quality judgment first vs. talent judgment first) to activate different naive theories more subtly. In both cases, participants only inferred high quality from high effort when an "effort" theory was rendered accessible, but not when a "talent" theory was rendered accessible. We conclude that judgment tasks prime naive theories that can serve as inference rules, illustrating that heuristics can be constructed on the spot.