Demand planning is a challenging component for organizations across a broad spectrum of industries. A key element of a successful demand plan is accurate forecasting, due in part to the operational decisions that are made based on the results of forecasting models. This is what our capstone project sponsor, Agility, has come to realize during their time-sensitive operations. Agility supplies food rations to the United Nations (UN) peacekeeping missions around the world. One particular mission, the mission in Darfur, or “UNAMID”, has a unique problem related to inaccurate forecasting. UNAMID places orders for food rations approximately 80 days in advance of when they are needed, but the lead time for Agility to source and deliver these items often exceeds 150 days. Therefore, forecasting is required to ensure that food rations can be procured and delivered on time. Currently, Agility uses a simple three-period moving average forecasting method, also known as MA(3). Due to frequent errors in the order quantity forecasted using this method, Agility often incurs stiff penalties from the UN for delivering too little or too much food. This study explores how sophisticated forecasting techniques can be applied to reduce penalty costs. First, we segmented the historical order quantity data to isolate the most important SKUs. Second, we tested various forecasting methods against the currently used MA(3) to determine if a more sophisticated model would produce better results. Third, we applied the Holt-Winters Forecasting model and optimized the parameters using non-linear optimization to maximize statistical accuracy. Fourth, we added penalty costs to the model and re-optimized the parameters to minimize the projected penalty cost. Fifth, we provided a set of strategic recommendations for how Agility can use the results of this study to realize these cost savings. We found that by using our optimized Holt-Winters forecasting model, Agility could likely save at least $25,000 per year in penalty costs at UNAMID. An additional study is recommended to explore how this model can be applied to further increase cost savings at other UN peacekeeping missions.