Do birds of a feather really sing together? Musicians face two competing pressures in the pursuit of success: conforming to genre norms to meet audience expectations and distinguishing themselves to attract the attention of listeners. These opposing logics may shape how artists choose their collaborators. Partnering with similar artists can reinforce genre alignment, while working with dissimilar collaborators may help artists stand out. This paper explores how these dynamics play out through the lens of homophily in musical collaboration. Drawing on network analysis, we develop a framework for measuring cultural proximity using song-level and collaboration data sourced from Spotify's API. Focusing on the understudied genre of Grime, we investigate whether the pull of similarity-the tendency to form homophilous ties-overrides competitive pressures within the genre.