Predicting factors that influence late delivery of sports apparel products in the supply chain
Loading...
Date
2024
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of the Western Cape
Abstract
The study aims to determine the factors that are related to sports apparel products being delivered late. The identification of these factors will assist supply chain management (SCM) and decision-makers (DM) with an understanding of the reasons behind the late delivery of sports apparel products.
Recent work in the field has shown that the most significant factor in predicting late delivery risk, using a binomial logistic regression model, was shipping mode. However, no regularisation techniques were used due to the pre-existing significant correlation between the variables. Machine learning methods are potential modelling approaches that could be used to overcome this. Previous work has also failed to address the existing factors such as the point of origin, order destination and how this may affect the late delivery of sports apparel products. The value in understanding the contributing factors to the late delivery of the sports apparel products can assist in improving timely deliveries, improve customer retention, which in turn can increase profitability. Furthermore, if the identification of these can assist in the timeously delivery of products, the company can be associated by customers as a reputable company that delivers on time thereby increasing customer loyalty. This study will made of various big data analysis and machine learning methods to build predictive models that will assist in the identification of these factors. The decision tree model exhibited the best fit, achieving a misclassification rate of 28.49% for training data and 28.67% for validation data, along with a receiver operating characteristic (ROC) index of 0.771 for training and 0.768 for validation data. They key shipping mode results indicated first and second class to be the worst performing whereas standard and same- day shipping produced better results in non-late delivery. The study identified a significant gap in the literature regarding the impact of shipping modes, order origin, and destination on the late delivery of sports apparel products. The findings provide insights for businesses, logistics providers, and policymakers aiming to optimise their SCM. By focusing on high-risk regions and seasons, stakeholders can anticipate and mitigate late delivery issues, improving operational efficiency and customer satisfaction. The limitation is the focus on sports apparel meaning the findings may not apply to other types of apparel or industries facing delivery delays. Addressing this limitation using empirical data could improve the analysis in future studies.
Description
Keywords
supply chain management, e-logistics, predictive analytics, decision-making