When an ant colony emigrates, its choice of a new home depends heavily on information sharing among the workers who organize the move. Behavioral heterogeneity is critical to this process, with some workers actively scouting new sites and recruiting nestmates, and others playing a more passive role. We have used machine intelligence and network analyses to refine our understanding of this heterogeneity, identifying four behavioral castes of workers—primary, secondary, passive and wandering—covering distinct roles in the spread of information during an emigration. This detailed characterization of the contribution of each worker can improve models of collective decision-making in this species and promises a deeper understanding of behavioral variation at the colony level.