In this thesis we address various factors that contribute both theoretically and practically to mitigating supply demand mismatches. The thesis is composed of three chapters, where each chapter is an independent scientific paper. In the first paper, we develop a framework that enables forecasters to use quantitative methods to forecast sales of newly launched products with limited historical data. The forecasts are provided over the new product's life cycle by leveraging data from similar older products. Our framework allows us to test autoregressive integrated moving average with exogenous variables (ARIMAX) and machine learning (ML) methods. It does so by exploiting the historical data of similar products through time-series clustering and by performing data augmentation to generate sufficient sales data. We perform a comparative analysis between ARIMAX and three deep neural networks (DNNs). Based on the results of using two data sets, we show that ARIMAX outperforms the more advanced DNNs. However, when adding Gaussian white noise to test the robustness of the methods, DNNs show better performance as the ARIMAX performance deteriorates with an increase in noise level. We provide insights for practitioners on when to use advanced forecasting methods and when to use traditional methods. The right choice of forecasting method leads to an increase in forecasting accuracy which results in better matches between supply and demand.

In the second paper, we study a price-setting newsvendor problem where the decision maker sells multiple products with uncertain demand and is faced with a capacity constraint on the total order quantity. To solve the problem, we first develop analytical expressions that make it possible to derive the optimal order quantities and optimal prices given a capacity limit. We show that the optimal allocation policy for the capacity is a nested-allocation policy. We show that the more limited the capacity, the more convergent the pricing policy. Optimal pricing policies change depending on the demand uncertainty and price elasticity. We therefore develop a decision typology that can be adopted by practitioners to choose the optimal pricing policy. It is important to study pricing and inventory simultaneously. Prices affect demand and demand affects order quantities. Therefore, optimizing both order quantities and prices has a positive impact on inventory management and profit.

Finally, in the third paper, we analyze the benefits of operational flexibility on both profitability and environmental sustainability. We measure the impact on the environment by the reduction in excess inventory at the end of the season when a certain operational-flexibility strategy is employed. We consider three different operational flexibility strategies: (1) lead-time reduction, (2) quantity-flexibility contracts, and (3) multiple sourcing. We find that the lead-time reduction strategy has the maximum capability to reduce waste in the sourcing process, followed by the quantity flexibility and multiple-sourcing strategies, respectively. Our results show that a firm can improve on both environmental sustainability and profitability by using an operational-flexibility strategy that localizes production and therefore provides better matching supply to demand.