| Forecasts are a critical input that drive actions within the firm and throughout the supply chain. For good reason, there is a tremendous focus on accuracy for this input. This dissertation addresses three areas regarding forecast accuracy in logistics and the supply chain relating to three questions posed by demand planners at a logistics provider firm that partnered with this research. In attempting to determine "What is causing our replenishment forecast error?", "What predictive factors can help improve our demand forecast accuracy?", and with regards to forecast accuracy "How good is good enough?", we explore three interrelated topics that have a broader impact on the academic conception of forecast accuracy than the original questions posed.;In three essays, we identify governance form factors that affect replenishment forecast deviation and bias, demonstrate accuracy improvement though the inclusion of uncertain weather forecast information in demand forecasts, and identify themes that serve to bound achievable and desirable demand forecast accuracy through a systematic literature review of logistics and supply chain journals. Our first study measures the deviation and bias related to franchise governance form, but also demonstrates a novel approach to contextualize the heterogeneity of effects across regionally, temporally and product category related conditions. Our second study expands on previous work linking the inclusion of uncertain weather forecast variables to improvements in demand forecast accuracy by examining a wider range of products and locations in a new industry, but also by demonstrating the limits to the value of uncertain information. Finally, our systematic literature review comprehensively presents the current state of research on the thematic drivers of forecast accuracy.;Each essay expands theoretical understanding of management phenomena, and reframes the manner in which previous research can be applied in practice. In each we also propose future avenues to expand on the work here, and on forecasting in general in the context of logistic and the supply chain. |
