Crowd-based innovation is used to tackle the complex and highly uncertain nature of innovation. It has been guided by the prevailing notion of parallel search, acknowledging the importance of crowd diversity and individual independence. Although using massive parallel paths to address uncertainty can help, it falls short of pursuing a well-organized crowd for innovation. Our parallelism abandons coordination potential too soon as parallel paths never interact. In response, we propose the notion of faceted search. With imperfect knowledge, individuals may mis-perceive the problem and focus on reduced facets. It re-opens the coordination opportunities in that different search facets can resonate with each other. Crowd wisdom is thus not simply the sum of parallel efforts, but naturally emerges from the interaction of individuals' wisdoms. We propose a tension in coordination between joint confirmation and mutual deviation to specify such resonance. We provide a concise and transparent computational model for scholars to advance the proposed faceted search theory. This paper contributes to the emerging literature on loose organizationality and offers practical implications for crowd-based innovation.
