Here, we demonstrate that the decision to conform to another person’s memory involves a strategic trade-off that balances the accuracy of one’s own memory against that of another person. We showed participants three household scenes, one for 30 s, one for 60 s, and one for 120 s. Half were told that they would encode each scene for half as long as their virtual partner, and half were told that they would encode each scene for twice as long as their virtual partner. On a subsequent two-alternative-forced choice (2AFC) memory test, the simulated answer of the partner (accurate, errant, or no response) was shown before participants responded. Conformity to the partner’s responses was significantly enhanced for the 30-s versus the 60- and 120-s scenes. This pattern, however, was present only in the group who believed that they had encoded each scene for half as long as their partner, even though the short-duration scene had the lowest baseline 2AFC accuracy in both groups and was also subjectively rated as the least memorable by both groups. Our reliance on other people’s memory is therefore dynamically and strategically adjusted according to knowledge of the conditions under which we and other people have acquired different memories.