180 likes | 233 Views
Learn how SkinCeuticals utilized the Modified Time Series Decomposition Model to improve accuracy in demand forecasting, reducing inventory costs, backorders, and lost sales. This presentation covers the company's history, the forecasting problem, methods of attack, the model's components, detailed results, and managerial recommendations for implementation.
E N D
Modified Time Series Decomposition Model The Weathermen: Rob Harrison RD Trinidad Niall Needham
Presentation Overview • Company Background • The Problem • Methods of Attack • The Model • Results • Managerial Recommendations
SkinCeuticals’ History • Founded in 1994 • Originally manufactured and sold 1 product • Now manufacture and distribute around 40 products to more to all 50 states and 30 countries worldwide
The Problem • Cash is being poured into stocking a large inventory to prevent backorders and potential lost sales • SkinCeuticals needs an accurate demand forecast for each individual product • Lead times
Methods of Attack • Associative Time Series? • Product clustering • Moving Averages • Forecasting Software • Time Series Decomposition • Modified Time Series Decomposition
Time Series Decomposition Model • Breaks forecast into four components • Trend • Seasonal • Cyclical • Irregular
The Components • Trend • b = (t × Yt – (t × Yt)/n) (t2 – (t)2/n) • a = Ŷ – b × ť • Cyclical • C = CMA Tc • Seasonal and Irregular • S × I = (T × C × S × I) (T × C) • Forecast • Ft = Tt × St × Ct
Modified Time Series Decomposition • Monthly Moving Error Forecast • Sales information for last year of sales deleted • Each of these months are forecasted • Forecast compared to actual sales data • Percent error is recorded and used in the final forecast
Managerial Recommendations • The Modified Time Series Decomposition Model will increase have greater forecasting accuracy with time • In the meantime, the Modified Time Series Decomposition Model forecast should be used along with associative forecasting techniques to create a more accurate demand forecast.