A novel approach is proposed for the development of a distributed simulation that is oriented towards the modeling and implementation of complex operational processes in the simulated world. We claim that, in the context of developing simulative systems, the proposed approach, Process Oriented Development (POD), bears significant advantages over "traditional" approaches for software development; specifically, we show that this approach allows for a significantly more effective quantitative analysis (i.e. the analysis of data collected in the simulation). The approach was employed in the context of military large-scale simulation experiments, and was proven to shorten both development and data analysis time, as well as to result in more accurate and reliable quantitative analysis results. The paper describes the main principles of the approach, analyzes its faults and merits with respect to traditional development processes, and presents a case study indicating the usefulness of the approach.