Each week, Daniels is featuring a researcher who conducts meaningful research that impacts their field and the wider community. Learn more about their work in Q&As with the Daniels Research team and email them to nominate yourself or a colleague for a future Q&A.
Mingrui (Ray) Zhang holds a PhD in Business Administration (Information Systems) from the University of Washington, a Master of Arts in Economics from Columbia University and a Bachelor of Science in Mathematics and Economics from the University of Illinois at Urbana-Champaign. His research utilizes both econometrical and analytical methodologies to examine the impact of technological innovations across various market contexts. Specifically, his work investigates the effects of new technology in retail, the dynamics of strategic partnerships between online and offline retailers, and the influence of consumer perceptions on the online crowdfunding market.
What do you study, and how did you become interested in this area?
My research mainly focuses on the impact of new technologies on customer behaviors and company outcomes. Early on, I worked with my advisor on projects involving e-commerce and crowdfunding platforms, and I found it fascinating to dig into customer behavior using those datasets.
I’m especially interested in how platforms can nudge customers in the right direction, and how businesses can design better strategies around these technologies. Over time, my work has grown to include both empirical analysis and modeling. That combination allows me to study concrete business problems and develop theoretical insights into how markets and customers interact.
How do you conduct these studies?
On the modeling side, I use economic modeling to explore dynamics between firms and customers. For example, I studied what happens when Kohl’s partners with Amazon to accept returns. We found there’s a small “win-win” region where both parties benefit. But when they do, it often comes at the expense of consumers, who face reduced surplus because firms can raise prices.
On the empirical side, I study crowdfunding platforms and supply chain finance. One project looks at how fundraisers position their ideas on Kickstarter to maximize success. Another examines guarantors in crowdfunded supply chain finance and how interest rates and credibility affect outcomes.
My dissertation examined smart vending machines in China. These machines, which let customers scan a code, open the door, and take what they want, create incredibly rich datasets. To utilize tensor completion algorithms, I built a three-dimensional tensor where the three dimensions represent region, scene(such as schools, hospitals, and hotels), and time, with transaction amounts as the tensor entries. This allowed me to estimate sales, fill missing values and predict how new placements—like putting a machine in a hospital in a certain region—might perform.
What are the practical implications of your work?
The vending machine project shows how companies can use high-dimensional data to improve inventory management and site selection. Even when some data is missing, the methods I use allow us to predict performance and identify new opportunities. For example, if a company doesn’t yet have a smart vending machine in a hospital, we can predict likely sales if they install one there.
More broadly, I think my research highlights the tradeoffs of new technologies. Firms may benefit from partnerships or innovations, but sometimes those gains come at the expense of consumers. By surfacing these dynamics, I hope businesses and policymakers can make more informed decisions.
How do you bring this research into the classroom?
I teach first-year students, so I don’t dive into full research studies, but I always bring in the ideas. In my data management course, I introduce students to different modalities of data—text, images, audio, video—and show how they can be transformed into datasets. I also present pieces of my vending machine study so students can see how theory meets practice.
Lately, I’ve been exploring how to use sports data in class, especially baseball. Baseball generates rich, high-dimensional data on swings, pitches, and game outcomes, and Rockies’ Coors Field has this unique “Coors Effect” that affect both pitching an field running. I think building a classroom project around this effect could be both fun and instructive for students
How do you hope your research impacts the world?
Right now, I’m working to get more of my work out to the public. For instance, our study on Amazon returns has real potential to spark conversations about monopoly power and consumer welfare. Amazon touches nearly everyone’s daily life, so showing the hidden tradeoffs in their partnerships could raise awareness.
Looking ahead, I want to deepen my focus on causal inference using high-dimensional data and AI. One of my recent projects investigates large language model–powered telemarketing agents and their effects on sales. I see this as a natural extension of my interest in how new technologies influence customers and firms, and I hope to build a body of research that helps people understand both the opportunities and the risks of these tools.