Optimizing Demand Planning Process - Seeking the Best Statistical Forecasting Method
Kuusirinne, Jussi (2014)
Kuusirinne, Jussi
2014
Konetekniikan koulutusohjelma
Talouden ja rakentamisen tiedekunta - Faculty of Business and Built Environment
Teknisten tieteiden tiedekunta - Faculty of Engineering Sciences
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2014-02-05
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tty-201401071012
https://urn.fi/URN:NBN:fi:tty-201401071012
Tiivistelmä
Companies are increasingly looking to plan their business more cost-effectively. Statistical forecasting has been deeply involved in the planning process to form a fact-based planning process. The purpose of statistical forecasting is to maintain and improve the current level of service, which affect the lead times. Forecasting allows companies to coordinate their production so that delivery times meet the demand. The case company has begun to use the Sales and Operations Planning process to help improve the company's business. The purpose of this thesis is to develop the above-mentioned process of finding the best possible statistical forecasting technique and categorize the case company's products in such way that they can be used for demand forecasting. The literature review will present several of statistical forecasting techniques and a technique which can be used to categorize the case company’s products. One of the purposes of the literature review is to give the reader a clear picture of how the supply chain, and Sales and Operations Planning are linked. The literature review works as a reference to find the best possible forecasting method for the case company’s data. Additionally, the literature review presents a categorization technique, which will categorize the case company’s products by their data’s variability and monetary importance. Based on aforementioned findings recommendations are made to aid Sales and Operations Planning Demand planning process.
The thesis discovered that the data analysis is particularly important. In the analysis it became clear how much the data have variation as well as whether there exist the seasonal variation in the data and trends. The results showed that the data had a seasonal variation and trends. According to the findings, after the comparison of the forecasting methods, the most suitable statistical forecasting method was found – Holt’s – Winters’. Based on the research, forecast accuracy can be improved by using X- categorized products having a low volatility. The study also looked for a correlation between the case company and the global industry indexes. According to the results, there was a positive or a negative trend with one month delay with a certain probability. The last task was to create a quality control for demand planning and forecasting. The purpose of the quality control was to follow the demand plan accuracy and to create control limits and also to find out if the most suitable forecasting method of the forecast tests is not too sophisticated for the time series. Conducting the long term performance analysis the historical demand plans need to be stored and compared with the actual figures. For this reason, a waterfall analysis was created. The purpose was to compare the current and past demand plans with actual demand figures. The key idea was to find out what is the error percent between the demand plans and the actual demand for an each given month in Sales & Operations Planning eighteen-month time frame. To this end, a solution was formed by a waterfall analysis which follows actual figures and forecasted values.
The thesis discovered that the data analysis is particularly important. In the analysis it became clear how much the data have variation as well as whether there exist the seasonal variation in the data and trends. The results showed that the data had a seasonal variation and trends. According to the findings, after the comparison of the forecasting methods, the most suitable statistical forecasting method was found – Holt’s – Winters’. Based on the research, forecast accuracy can be improved by using X- categorized products having a low volatility. The study also looked for a correlation between the case company and the global industry indexes. According to the results, there was a positive or a negative trend with one month delay with a certain probability. The last task was to create a quality control for demand planning and forecasting. The purpose of the quality control was to follow the demand plan accuracy and to create control limits and also to find out if the most suitable forecasting method of the forecast tests is not too sophisticated for the time series. Conducting the long term performance analysis the historical demand plans need to be stored and compared with the actual figures. For this reason, a waterfall analysis was created. The purpose was to compare the current and past demand plans with actual demand figures. The key idea was to find out what is the error percent between the demand plans and the actual demand for an each given month in Sales & Operations Planning eighteen-month time frame. To this end, a solution was formed by a waterfall analysis which follows actual figures and forecasted values.