Demand Forecasting For Bought-in Materials Using Time-Series Methods
Do, Vi Phung-Thao (2019)
Do, Vi Phung-Thao
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
This product-based thesis is created to support the commissioning company in predicting the future demand of bought-in materials using time-series forecasting methods. The company is a manufacturer of special-purpose machineries, in which demand forecasting has not been implemented earlier. The thesis objectives are to introduce suitable forecasting methods that fit the case item demand data, generate new forecasts for the upcoming periods and develop a simple tool that assists the company in the forecasting process.The project management methods used in the thesis include desk research, qualitative research and quantitative research. Desk research is used to develop a thesis topic-related theoretical framework. The qualitative method is implemented to obtain information regarding the needs for demand forecasting and identify suitable item cases through face-to-face interviews with the commissioning company. The quantitative method is primarily applied to collect and analyze numerical data, develop forecasting models and generate new forecasts using Microsoft Excel and IBM’s SPSS software.The theoretical framework of the thesis is developed using the desk research method with a focus on demand forecasting in supply chain management and time-series forecasting. It includes the definition and explanation of concepts, equations and techniques on how to analyze time-series data and utilize time-series forecasting approaches to conduct new forecasts.The empirical part of this thesis primarily explores the methods chosen for forecasting historical data that exhibit trend and seasonality patterns, namely Holt exponential smoothing and Holt-Winters exponential smoothing methods. The deliverables would be a goodness of-fit measurement between the forecasting models and historical data, new forecasts of year 2019 and an Excel-based user-friendly tool which can be used to forecast the future demand of similar case items.In the Discussion chapter, the key findings derived from the project tasks are explained, including the theoretical framework summary, the forecasting process outcomes and the Excel forecasting tool. The company’s feedback on the thesis results is then presented. Recommendations for further forecasting implementation in the company and forecast accuracy enhancement are also provided. Finally, the thesis report ends with an overall project assessment and a reflection of the thesis author’s own learning.