Abstract

In this dissertation, I leverage artificial intelligence (AI) for making effective marketing decisions, specifically on how to measure the causal effects of targeting promotions and detecting suspicious retail buyers of opioids. In the first chapter, we propose a novel approach in measuring the true causal effects of targeting promotions. Targeting promotions on online platforms are often determined by AI algorithms, which utilize extensive customer and seller information to generate various algorithmic scores for targeting. Effective targeting, however, will lead to selection bias when evaluating the causal effects of promotions. we analyzed 2,294 promotion experiments on a major online retail platform and found that traditional methods, such as propensity score matching and double machine learning, cannot accurately recover the true effects using readily available data. To overcome this challenge, we propose an approach that logs and utilizes the algorithmic scores to match treated and untreated customers, effectively mitigating the selection bias and addressing the curse of dimensionality in matching. We validate this approach by analyzing the same set of experiments and demonstrate that the estimates from the proposed matching approach, based on algorithmic scores, closely align with the promotion effects estimated from the separately run randomized field experiments. This approach can assist platforms and sellers in accurately evaluating the value of targeted promotions. Additionally, it can be implemented easily and at a low cost since algorithmic scores are easy to store. In the second chapter, we propose an anomaly detection algorithm to effectively detect suspicious opioid retail buyers. The opioid epidemic adversely affects thousands of communities across the US. The objective in this research is to examine how a US drug distributor can leverage cutting-edge analytics to monitor opioid diversion and stop opioid shipments from reaching those at risk. In doing so, the firm can lead with purpose and create maximum social impact through CSR. The authors propose an anomaly detection algorithm that can be used to identify suspicious retail buyers of opioids. The authors implement the proposed algorithm on the ARCOS database -- which tracks all opioid drug shipments across the US from 2006 to 2012. The proposed algorithm effectively identifies suspicious retail pharmacies and practitioners involved in drug diversion. It achieves 100 % precision and 100 % sensitivity, resulting in 100 % F-1 score for retail pharmacies. By applying the proposed algorithm, the drug distributor gains a powerful tool for promptly detecting suspicious retail buyers. By reporting suspicious opioid orders as they occur to the DEA, the distributor can safeguard vulnerable communities and save lives. Putting societal purpose before short-term profit will build sustained competitive advantage for the drug distributor in the long run.

Committee Chair

Tat Chan

Committee Members

Dennis Zhang; Meng Liu; Seethu Seetharaman; Zhenling Jiang

Degree

Doctor of Philosophy (PhD)

Author's Department

Marketing

Author's School

Olin Business School

Document Type

Dissertation

Date of Award

5-22-2025

Language

English (en)

Download

Included in

Marketing Commons

Share

COinS