Improving Demand Prediction and Reducing Out-of-Stock – Application of Advanced Data Analytics in Retail Supply Chains

Research output: Chapter in Book/Report/Conference proceedingsConference contributionpeer-review

Abstract

Correct demand prediction is a key success factor of efficient and demand-driven
supply chains. This is especially true for the retail sector, where out-of-stock products directly influence customer satisfaction. In a data-driven world, advanced analytics approaches offer huge potential for demand prediction. The paper applies two of the most acknowledged demand prediction approaches to a real-world retail case. Based on an extensive database of cashier as well as Supply Chain data, we apply ARIMA and SARIMA and evaluate their applicability to predict the demand of selected perishable products. In addition, the impact of adding SARIMA-based demand forecasting to out-of-stock detection is analysed. The results show high
applicability and a good forecasting quality especially of SARIMA. The quality of out-of-stock detection can significantly be improved by adding advanced analytics to traditional approaches in this area. For reaching higher demand prediction quality, results indicate the need to add the effect of promotions and the implications of substitute products to the applied approach.
Original languageEnglish
Title of host publicationProceedings of the 8th International Physical Internet Conference
Number of pages9
Publication statusPublished - 2021

Cite this