Selected Abstracts


Consumer Engagement with Sequential Content: A Content-Aware Dynamic Choice Model (with K. Sudhir and Tong Wang)

Sequential content—such as episodic fiction, television series, and online courses—requires consumers to make sequential decisions about whether to continue by incurring access costs or to exit. These choices reflect forward-looking tradeoffs among engagement, delay, and payment, and depend jointly on platform policies and the content encountered along the progression. Yet incorporating high-dimensional content into dynamic choice models remains challenging. We address this by extending adversarial inverse reinforcement learning (AIRL) to accommodate unobserved heterogeneity, identifiable action-dependent utilities, and scalable representations of unstructured content within a structural dynamic framework. We apply the model to 15 months of detailed readership and payment data from a serialized fiction platform. The estimates reveal two dominant segments—Pay & Read and Wait & Read—that account for most consumption and nearly all purchases. Counterfactual analyses show that free access can increase downstream purchases within a user, delay policies can shift engagement toward paid consumption, and content-aware pricing improves monetization while preserving overall usage. More broadly, the paper provides a scalable approach for estimating dynamic choice models with high-dimensional state spaces, offering a tractable framework for studying sequential engagement in modern digital markets.




Pay Now or Wait to Unlock? Cliffhangers and Monetization of Serialized Media (with Vineet Kumar and K. Sudhir) 

Serialized media, including multi-episode books and shows, commonly uses a “wait-to-unlock” model wherein consumers can pay immediately for a new episode or wait for free access. We leverage a large-scale natural experiment on a U.S.-based fiction platform to examine the causal impact of different wait-time reductions across 10,000 series and over a million users. Employing a matching-based difference-in-differences framework, we show that shorter waits substantially increase aggregate consumption by both drawing in new readers \emph{and} boosting engagement among existing ones. However, effects on paid unlocks vary with the extent of wait reduction: moderate cuts can drive revenue gains, while drastic cuts can cannibalize paid consumption—yet in certain series, deep reductions also intensify paid use among loyal readers. Crucially, we find that these heterogeneous effects are related to the “sequential complementarity” (operationalized as cliffhanger strength) across consecutive episodes, i.e., how strongly each episode’s content drives immediate consumption of the next. We show that high complementarity mitigates cannibalization, leading consumers to purchase even when wait-times are drastically lowered. By highlighting how temporal “distance” between free and paid episodes interacts with sequential complementarity, our findings advance versioning theory and help platforms to devise book-specific wait to unlock strategies.



Artificial Intelligence Applications to Customer Feedback Research: A Review (with Ishita Chakraborty and Shrabastee Banerjee)

In this paper, we aim to provide a comprehensive overview of customer feedback literature, highlighting the burgeoning role of artificial intelligence (AI). Customer feedback has long been a valuable source of customer insights for businesses and market researchers. While previously survey focused, customer feedback in the digital age has evolved to be rich, interactive, multimodal, and virtually real time. Such explosion in feedback content has also been accompanied by a rapid development of AI and machine learning technologies that enable firms to understand and take advantage of these high-velocity data sources. Yet, some of the challenges with traditional surveys remain, such as self-selection concerns of who chooses to participate and what attributes they give feedback on. In addition, these new feedback channels face other unique challenges like review manipulation and herding effects due to their public and democratic nature. Thus, while the AI toolkit has revolutionized the area of customer feedback, extracting meaningful insights requires complementing it with the appropriate social science toolkit. We begin by touching upon conventional customer feedback research and chart its evolution through the years as the nature of available data and analysis tools develop. We conclude by providing recommendations for future questions that remain to be explored in this field.



Works in Progress