Two essays on managing the retail marketing mix during a product-harm crisis

Abstract

The retail sector has a long history. Studies of retail management strategies also make up a significant portion of the marketing literature. However, many important problems are yet to be answered. Managing the retail marketing mix during a product-harm crisis is one of them. This dissertation examines two outstanding issues that are related to this topic. The first essay studies the problem of weathering a product-harm crisis for retail brands. Previous research mainly focused on understanding consumer responses to a product-harm crisis in Western markets. In contrast, little is known about retailer reactions in such cases or crisis impacts in other parts of the world. This essay demonstrates the vital role of retail distribution to a brand during the crisis. The study investigates the impact of the 2008 infant formula scandal in China with a Multivariate Dynamic Hierarchical Linear Model. Results confirm the crucial effect of distribution on brand sales during a crisis, highlighting the role of retailers in these situations. In line with previous findings in Western markets, pricing or advertising strategies are likely ineffective to influence consumers due to the crisis. Yet, this study finds that the brand’s product line length has significant effects on its distribution and the subsequent sales outcomes. The second essay makes a significant methodological contribution to solve a missing data problem in pricing data that has encumbered empirical retail research for decades. Even though the scope and richness of retail datasets have grown rapidly in recent years, most of them only record actual prices paid by customers and lack direct information about regular prices and discounts. This lack of data limits the insights researchers could generate since consumer sensitivities to regular price and discount changes typically differ.

This essay introduces a new machine learning algorithm that decomposes prices using Atheoretical Regression Trees (DEPART). Our benchmarking shows that the proposed method systematically outperforms previous approaches by about fifty percent or more in terms of accuracy, which can help firms improve their pricing decisions.

Date of Award	2022
Original language	English
Awarding Institution	The Hong Kong University of Science and Technology
Supervisor	Ralf VAN DER LANS (Supervisor) & Kristiaan HELSEN (Supervisor)