After a movie achieves box office success, it may be broadcast as a television drama or even released as a computer game. What, then, is the difference between a movie being broadcast as a drama versus being produced as a computer game? In terms of perceptual distance, movies and dramas are very close, whereas the distance between movies and computer games is far. That is, extension from a movie to a drama is a close extension, while extension from a movie to a computer game is a distant extension. According to cohesiveness theory, when information is arranged in an organized manner that facilitates easy comparison, cohesiveness is high. Conversely, when information is arranged in a scattered and disorganized manner that makes interpretation difficult, cohesiveness is low. Therefore, extension from a movie to a drama is a high-cohesiveness extension, while extension from a movie to a computer game is a low-cohesiveness extension. Should a high-cohesiveness extension such as a drama be similar to the original movie, or should it be differentiated? Conversely, should a low-cohesiveness extension such as a computer game be similar to the original movie, or should it be differentiated? Through three experiments, this study identified the situations in which similarity to the original is beneficial and those in which it is not. Experiment 1, examining extensions reflecting cohesiveness (e.g., producing a drama based on the movie "Silmido" vs. producing a computer game based on the movie "Silmido"), found that for high-cohesiveness extensions, consumer responses became more positive as differentiation from the original movie increased. Similarly, for low-cohesiveness extensions, consumer responses also became more positive as differentiation from the original movie increased. Experiment 2 investigated the effect of involvement with the movie. The results showed that for consumers with low involvement, the cohesiveness effect appeared in movie brand extensions. However, for consumers with high involvement, the cohesiveness effect did not appear. This is because when involvement is high, consumers focus more on the substantive content rather than contextual effects such as information cohesiveness. Experiment 3 investigated the effect of need for uniqueness (NFU). The results showed that consumers with high NFU exhibited results opposite to the cohesiveness effect. Consumers with high NFU preferred character consistency. However, in low-cohesiveness extensions, they preferred character change.