Research
I have two primary research objectives. First, I aim to generate work that holds significance both in academic and in practical business contexts. Second, I strive to produce research that is accessible and comprehensible to a broader audience, including those outside the field.
I am currently focusing my research on the domains of non-fungible tokens and dating. My papers have been invited for presentation at various premier venues and have received nominations for best paper (ICIS 2023) and best student paper (WISE 2022). My research pipeline includes further exploration of non-fungible tokens and dating, as well as expansion into the realms of artificial intelligence, online streaming, and real estate.
I primarily acquire data for my research through methods such as web scraping and accessing APIs. Additionally, in some cases, I am also incorporating experimental research methods and collaborating with companies.
Journal Publications
NFT Disruption in Platform Competition: Evidence from Trading Card Collectibles
(Joint work with Dominik Gutt and Ting Li)
Information Systems Research
Work in Progress
When Creation Is Cheap but Attention Is Not: Congestion Dynamics in Creator Markets
(Joint work with Dominik Gutt, Murat Tunc, and Ting Li)
Current stage: Under review at Management Science
Abstract: Reducing creators' marginal costs is a common platform strategy to expand content supply. Yet platform success depends not only on producing more content, but also on allocating scarce user attention to it. In this study, we examine the engagement and market consequences of cost-side scaling in creator-driven marketplaces by exploiting the introduction of lazy minting on a leading NFT platform - an exogenous reduction in creation costs that removes upfront minting fees. Using a difference-in-differences design comparing it to similar NFT platforms that did not adopt the policy, we find that lower creation costs sharply expand token supply but reduce engagement with new tokens. Post-policy tokens receive fewer likes, are less likely to receive any engagement, and, conditional on engagement, attract fewer interactions. The engagement decline is not explained by changes in observable creator effort or by compositional shifts in who creates tokens. Instead, it is larger when token congestion is high and is most pronounced shortly after launch, consistent with weaker discoverability and heightened cold-start frictions when supply grows faster than attention. The decline is driven mainly by reduced non-follower attention and is most severe for tokens that fail to gain early traction, indicating greater path dependence in attention allocation. These dynamics have economic consequences: post-policy tokens are less likely to sell, while first-sale prices are largely unchanged. Overall, our findings highlight a tension in digital creator marketplaces: lowering entry costs can expand supply but may weaken engagement and market outcomes when attention and discovery do not scale proportionally.
How Anti-ghosting Interventions Affect Matchmaking
(Joint work with Rodrigo Belo and Ting Li)
Current stage: Preparing journal submission
Abstract: Many digital platforms do good well in bringing users together but not so much in encouraging continued communication. In this work, we collaborate with a dating platform and study how an "anti-ghosting" feature aimed at continuing communication affects users. We find that users exposed to this feature increase all their communication activity, even that which is not directly associated with this feature. We study why this happens and find that exposed users browse less for other members and instead converge their attention towards members they are already connected to. Studying matchmaking, we find that exposed users benefit both by becoming more likely to move their communication off the platform with other members, as well as by doing so with more members. Furthermore, we assess the economic value of incorporating this feature for platforms. Lastly, we discuss how we contribute to the literature and our work's implications for platform decision makers.