Good supply strategy that can improve the logistical chain involves taking into account the complexity that a specific good is situated in. This research takes improving the bicycle supply chain within the Netherlands armed Forces as a use case. Aside from the goals for the use case of bicycles, the aim was also to fill the knowledge gap of analyzing an intra-organizational, multi-structural supply chain system, from the perspective of viewing an organization as a collection of business processes. The method used to fill the knowledge gap and gain an answer on the research question is Agent-Based Modelling, combined with an exploration approach using the Exploratory Modelling and Analysis (EMA) workbench. The results show that If the processes are not aligned at every decision made by agents, extra variance and unpredictability in the behavior is seen. Both predictive replacements and willingness to share information were promising policies, but do require a trade-off. The uncertainty in demand and supply also greatly influenced the effort required of the different agents. Choices have to be made in what found important within the supply chain. One must also take into account the long term visions of Defense with regards to the goals of the organization as a whole. The used simulation methodology in this research is a novel approach to investigate complex but practical problems occurring within Defense’s processes. Recommendations to overcome hurdles in practical research are, establishing a clear pragmatic goal and exploring all the processes involved before implementing a simulation model or looking to explain quantified performance indicators. Analyzing the intra-organizational supply chain within Defense showed that in general, intra-organizational supply chain research would benefit from a more process oriented perspective. Generalizing of the results goes as far as goods that have the same classification. Limitations of this research consist of three points: Effort as KPI, run time and extrapolation. Further research might investigate the possibility of extrapolation, extended geographical structure within the model and expansion of uncertainty considerations.