Essays on Modeling Choices in Experiential Goods Categories
Ahn, Gwen
2023
Abstract
My dissertation explores how firms can use individual longitudinal choice data to increase usage and spending in experiential goods categories, specifically in two different live events domains: art performances and sports events. In the first essay, I study how consumers choose customizable bundles of art performances while balancing preferences for the constituent items and within-bundle variety. Known as the diversification effect, consumers prefer higher levels of variety when making choices for multiple consumption occasions. This suggests that picking the right level of variety is a crucial part of art performance bundle choices. While customizability adds flexibility to consumers’ choices, due to the large number of possible bundles, it could increase consumers’ cognitive costs. Based on the proposed model, I make individualized recommendations that reflect heterogeneous preferences for not only the performances but also for the level of variety. To model consumers’ choices of art performance bundles, it is necessary to 1) characterize the performances, 2) define and measure the variety of prospective bundles, and 3) devise an efficient way of estimating bundle choices in the prohibitively large combinatorial space of possible bundles. Using natural language processing, I extract latent dimensions of the art performances that can be used to characterize the performances. I use these latent attributes to construct variety metrics that capture consumers’ perception of bundle-level variety. Additionally, I devise a novel Monte Carlo approach to integrate over the space of unobserved order in which the bundle was assembled to tame the curse-of-dimensionality in the estimation process. I find that including variety metrics substantially improves predictive performance of the model, allowing the performing arts organization to make better individualized recommendations. In the second essay, I study the cross-channel structure of the National Football League (NFL) ticket markets and consumers’ purchase channel choices with the goal of devising optimal dynamic pricing and inventory policies across different channels. Professional sports teams have widely adopted dynamic pricing policy, which resulted in significant revenue improvements. At the same time, the growth of legal secondary markets has contributed to the development of a complex market structure with multiple channels. Understanding the cross-channel structure and consumers’ channel choice process allows teams to make more informed pricing and inventory decisions. Partnering with an anonymous NFL team, I collected time-series data on the availability and pricing of tickets on primary and three major secondary channels and combine it with transaction data. I propose a three-part model to understand the supply and demand dynamics: sellers’ supply decision, buyers’ purchase decision, and channel choice to capture the evolution of the choice environment where ticket availability and prices vary over time and channels. I find that there exist significant price differences across channels even after controlling for seat quality, and that channel choices reveal differential price sensitivities, effects of time-until-game across channels, and strong past dependence. Importantly, the investigation into row-level supply decisions reveal potential cross-channel effect of sudden increase in ticket availability on the primary channel due to an unexpected buyback from the brokers, opening a window to investigate the causal effect of supply changes across the channels.Deep Blue DOI
Subjects
Choice models Experiential consumption Variety-seeking Dynamic pricing Live events
Types
Thesis
Metadata
Show full item recordCollections
Remediation of Harmful Language
The University of Michigan Library aims to describe library materials in a way that respects the people and communities who create, use, and are represented in our collections. Report harmful or offensive language in catalog records, finding aids, or elsewhere in our collections anonymously through our metadata feedback form. More information at Remediation of Harmful Language.
Accessibility
If you are unable to use this file in its current format, please select the Contact Us link and we can modify it to make it more accessible to you.