Analysing the role of information exchange for demand forecasting in collaborative supply chains.
PhD thesis, University of Nottingham.
It is now widely recognized that supply chains, not individual organisations, are responsible for the success or failure of businesses. This has necessitated close coordination among supply chain partners. In the past few decades, in an attempt to improve the overall efficiency of the supply chain, many companies have engaged in collaboration with other supply chain members. Consequently, several supply chain management initiatives such as Vendor Managed Inventory, Efficient Consumer Response, Continuous Replenishment and Accurate Response have been proposed in the literature to improve the flow of materials as well as information among supply chain partners. In this line, Collaborative Planning Forecasting and Replenishment (CPFR) is a relatively new initiative that combines the intelligence of multiple trading partners in planning and fulfilment of customer demand by linking sales and marketing best practices. The role of CPFR has been widely studied in the US retail industry, but it has not been researched much in the UK and also in Asian countries. Hence, this research focuses on the adoption of CPFR in the UK and India.
Levels of collaboration and information sharing differ to a great extent across the supply chains based on the needs of individual businesses. Accordingly, the importance of CPFR varies in different supply chains. The study reported in this research explores the operations of CPFR and highlights the corresponding benefits in different firms using case studies of Indian (4 cases) and British (2 cases) companies operating in Make-To-Stock (MTS) and Make-To-Order (MTO) environments. In this research, information exchange among collaborating partners is analysed with a focus on its role in demand forecasting and timely replenishment.
In order to identify potential benefits of CPFR, this research has adopted a four stage approach. In the first stage, interviews with top and middle managers in the case companies helped to develop a clear understanding of the collaborative arrangements in each company. In stage two, a conceptual model called the Reference Demand Model (RDM) was developed. RDM is a specific model representing the dependency of demand projection on information from different supply chain members involved in supply chain processes. When fully developed, the RDM will serve as a decision tool for the companies involved in collaboration to decide on the level of collaboration and the type of information exchange in order to improve supply chain planning and forecasting.
Further, to explore how demand information collected through RDM can help improve forecasts accuracy, a quantitative approach is employed in the next two stages. Therefore, stages 3 and 4 were studied only for the cases with detailed sales data. In stage 3, structural equation models were developed to establish the underlying relationships among demand factors that were identified using RDM. In stage 4, regression forecast models of sales were developed using the demand factors identified through RDM. The forecast models showed an improved accuracy and thus this research suggested the case company (Soft Drink Co.) to use the demand information (identified from RDM) in the demand forecasts.
The results strongly support CPFR in a MTS environment with promotional sales, and exchanging the detailed sales information from downstream to upstream supply chain members may improve the accuracy of demand forecasts. Information exchange is also required to ensure timely replenishment for MTS products. However, in a MTO environment, there is less need for collaboration with downstream supply chain partners for the purpose of short term demand forecasting.
Thesis (University of Nottingham only)
||H Social sciences > HD Industries. Land use. Labor
||UK Campuses > Faculty of Social Sciences, Law and Education > Nottingham University Business School
||25 May 2011 10:50
||13 Sep 2016 12:39
Actions (Archive Staff Only)