Do hierarchies aid adaptation? When do they do so better than crowds of autonomous agents or communities of peers? Using computational models, we contrast three distinct organizational forms --- a branching hierarchy, symmetric “community” like structures, and “crowds” of unconnected actors. Abstracting from issues of incentives and goal conflict, we focus on how these alternative forms are adaptive with respect to their role in influencing beliefs. The superiority of one form over another is shown to depend on the nature of the task environment, as represented by the difficulty of the search problem and the degree to which coordination among components of organizational action is important, and the characteristics of the individual actors, as represented by their rate of learning and their proclivity to engage in exploration. Alternative structures influence beliefs differently based on the degree to which influence is symmetric or asymmetric. Relative to no influence (crowd), symmetric influence processes (communities) are a prod to exploration, while asymmetric influence (hierarchies) boosts convergence in actions. These contrasting forces, in conjunction with the properties of the individual actors, map out distinct settings of superior organizational forms.