Challenges in Retail Supply Chains: Strategies for Optimizing Logistics and Enhancing On-Shelf Availability

This thesis begins with a systematic literature review of the prevalent theories in supply chain management and retailing journals, specifically those related to on-shelf availability (OSA) and out-of-stock (OOS) studies. It highlights that Service-Dominant logic and Inventory Theory are the two key theories addressing both pre-store and in-store issues. Additionally, it points out that deductive reasoning is more commonly applied than inductive reasoning in these studies. The second essay explores OSA and the prevention of OOS situations, focusing on the operational aspects of managing OSA and OOS. Furthermore, there is a need to investigate the behavioral factors influencing decision-making processes across various departments within a retail supply chain (RSC) from the cross-functional integration (CFI) perspective. By conducting a case study, this essay emphasizes the decision-making behaviors of managers interacting across different business functions in an internal retail supply chain. It unpacks the mechanisms and boundary conditions of CFI, providing valuable insights for retailers aiming to improve OSA and OOS performance, particularly in the apparel industry. The third essay delves into the complexities of the apparel industry, where the diverse range of products makes accurate volume estimation challenging. Inaccurate volume predictions can negatively impact decision-making in logistics planning, increasing costs and reducing profitability. While existing literature has explored transport planning based on historical order sizes, research on AI applications in forecasting, logistics optimization, and real-time tracking is still in its early stages. This study employs Design Science (DS) providing a framework for addressing complex problems, benefiting both academics and practitioners. It compares various well-known machine learning (ML) algorithms with the current calculation methods used by a global retailer. The contributions of this research offer new theoretical and managerial insights, particularly in the domains of Logistics and Supply Chain Management (L&SCM), with a focus on procedures, metrics, and model selection.

This thesis begins with a systematic literature review of the prevalent theories in supply chain management and retailing journals, specifically those related to on-shelf availability (OSA) and out-of-stock (OOS) studies. It highlights that Service-Dominant logic and Inventory Theory are the two key theories addressing both pre-store and in-store issues. Additionally, it points out that deductive reasoning is more commonly applied than inductive reasoning in these studies. The second essay explores OSA and the prevention of OOS situations, focusing on the operational aspects of managing OSA and OOS. Furthermore, there is a need to investigate the behavioral factors influencing decision-making processes across various departments within a retail supply chain (RSC) from the cross-functional integration (CFI) perspective. By conducting a case study, this essay emphasizes the decision-making behaviors of managers interacting across different business functions in an internal retail supply chain. It unpacks the mechanisms and boundary conditions of CFI, providing valuable insights for retailers aiming to improve OSA and OOS performance, particularly in the apparel industry. The third essay delves into the complexities of the apparel industry, where the diverse range of products makes accurate volume estimation challenging. Inaccurate volume predictions can negatively impact decision-making in logistics planning, increasing costs and reducing profitability. While existing literature has explored transport planning based on historical order sizes, research on AI applications in forecasting, logistics optimization, and real-time tracking is still in its early stages. This study employs Design Science (DS) providing a framework for addressing complex problems, benefiting both academics and practitioners. It compares various well-known machine learning (ML) algorithms with the current calculation methods used by a global retailer. The contributions of this research offer new theoretical and managerial insights, particularly in the domains of Logistics and Supply Chain Management (L&SCM), with a focus on procedures, metrics, and model selection.