Digital trace data present researchers with unique opportunities to investigate and theorize various socio-technical phenomena. The availability of advanced computational methods allows researchers to surface latent patterns from digital trace data. Computational approaches to theory construction have emerged as a new research genre that exploits these capabilities. Moving from computationally derived patterns to theory is a challenging process. Researchers can study multiple phenomena using digital trace data, use multiple computational methods, and apply multiple theoretical lenses. This abundance of possibilities introduces significant uncertainty in constructing meaningful theoretical contributions. Therefore, it is crucial to identify a pathway that ultimately leads to meaningful theoretical contributions. This chapter presents two pathways toward theory construction using digital trace data, priori theorizing and situated theorizing, along with a discussion on the challenges and opportunities associated with each pathway.