Intuitively, insight emerges unexpectedly. However, some previous views proposed that insight emerges with a high probability after people recognize their failure in solving a problem. In order to empirically investigate this failure-insight relationship, this study manipulated when participants recognized failure by using social comparison. It presumed that participants who had not yet solved the problem but knew others had already solved it would recognize that their currently adopted strategy was a failure; the timing of this was manipulated in the experiment. As expected, participants who were given a cover story regarding othersʼ fast performance for the T-puzzle completed the same puzzle more successfully, as compared to those who were given a story of othersʼ slow performance. The results suggest that the occurrence of insight was influenced by when participants recognized their failure. Providing social reality information (i.e., othersʼ good/poor performance) might be a method to facilitate or inhibit insightful problem solving.