Electronic Thesis and Dissertation Repository

Thesis Format

Integrated Article

Degree

Doctor of Philosophy

Program

Business

Supervisor

Pun, H.

2nd Supervisor

Ghamat, S.

Affiliation

Wilfrid Laurier University

Co-Supervisor

Abstract

I study irregular entries to the end-customer market and the impact of such entries on suppliers, buyers, and customers. I am particularly interested in the irregularities of supplier encroachment and counterfeiting problems. This dissertation addresses these issues and proposes solutions in the form of three essays. In the first essay, I study a supply chain, consisting of a supplier and a buyer where the supplier can encroach on the end-customer market and keeps private information on its own production capacity. The supplier can decide on its capacity allocation and the buyer can order strategically, hoarding the supply capacity, to remove the competition. I find that the supplier is worse off, and the buyer is better off, when the supplier keeps its capacity information private. Further, I demonstrate that the supplier may no longer encroach on the end-customer market when it has more capacity. The second and third essays are inspired by the counterfeiting problem on online e-commerce platforms. In the second essay, I develop an algorithm that analyzes customers’ reviews on an online platform and provides an authenticity score for the products. I trained context-specific word embedding based on a large corpus of Amazon customer reviews to show that my unsupervised methodology provides good predictive power. Next, I study the effect of customers’ reviews on an e-commerce platform’s anti-counterfeiting strategy against third-party sellers. The platform can provide a tool for customers that analyzes just the product reviews or a more advanced tool that analyzes both the product and seller reviews to help customers determine if products are fake or genuine. On the seller’s side, it can choose to reveal its fake products by charging a lower separating price based on its profit under these two options. I demonstrate that even when the tools are free, the platform does not provide the advanced tool if the seller sells products with a low authenticity score (fake products), and it provides the basic tool if and only if the demand of the genuine product is sufficiently high. Together, these papers provide solutions on how to maximize profits by making informed decisions in the face of market irregularities for the supplier, the buyer, and the customer.

Summary for Lay Audience

In this dissertation, I study abnormalities in the end-customer’s market and how they impact suppliers, buyers, and customers. In particular, I am interested in factory-direct selling and counterfeiting problems on online e-commerce platforms. In this dissertation, I study a situation where a supplier decides to sell directly to the customers and compete with its buying firm. Because the supplier has a limited amount of product, it must allocate it between its own selling channel and the buyer. Because the buyer does not know how much product the supplier has, the buyer may aggressively over-buy the product to control the competition in the end-customer market. I show that contrary to the common belief, keeping its product amount private hurts the supplier’s profit because it must signal its production capacity to the buyer. Moreover, having excess product hurts the profit of both the supplier and the buyer because, in this case, the firms have to compete in the end-customer market.

I also study counterfeiting problems on online e-commerce platforms. First, I use machine learning techniques in text analysis to develop a counterfeit detection algorithm. My algorithm analyzes previous customers’ reviews for indications of fake or genuine products on online platforms and provides an authenticity score for each product. I test the accuracy of my algorithm by manually assigning a label (i.e., fake/genuine) to the reviews using human coders. Second, I study the strategic behavior of online e-commerce platforms and third-party sellers with regard to using platform-provided anti-counterfeiting tools. I demonstrate that even when the tool is free, the platform does not provide an advanced counterfeiting tool if the seller sells products with a low authenticity score (fake products). Moreover, I show that the platform provides a basic counterfeiting tool if and only if the demand of the genuine product is sufficiently high.

Download

Share

COinS