Working Papers

The Impact of Bifurcation on Platform Outcomes in a Q&A Community (Job market paper)

with Chris Forman and Michael Kummer

This paper studies platform bifurcation, where a subgroup of users from the original platform launches an independent spin-off platform. We identify the effects of bifurcation using a DID approach, which exploits the introduction of spin-off platforms in an online platform incubator. We find that bifurcation leads to a strong overall increase in contributions. While contributions in the home platform decline, the two bifurcated platforms generate more combined user contribution and attract more new users compared to a single united platform. We further explore how interconnectivity and platform differentiation affect users’ platform choice. Our evidence indicates that users are less likely to migrate from an incumbent platform to a new specialized platform when the interconnectivity is strong. However, users are more likely to migrate when the specialized platform enables more differentiation. This paper is the first to empirically analyze the strategic implications of new platform entry at scale and to document the moderating role of interconnectivity and platform differentiation.

Chat More and Contribute Better: An Empirical Study of a Knowledge-Sharing Community

with Chris Forman and Michael Kummer

Under revision for second round review at Information System Research [SSRN]

Knowledge production platforms usually impose norms of knowledge contribution and employ gamification systems such as badges and reputation. Such designs may have unwanted outcomes such as inducing competition and creating entry barrier for new users. A secondary channel that allows users to interact informally may resolve those limitations in designs in the main platform. In this paper, We investigate the impact of opening a secondary channel in a knowledge production platform. We identify the causal effect of the secondary channel in a question and answer platform, Stack Overflow, by exploiting the introduction of the chat rooms and using a difference-in-differences approach. We have several findings: First, chat rooms improve efficiency of knowledge exchange in the main platform. The introduction of chat rooms increases the likelihood of receiving a satisfied answer. Second, chat rooms increase user’s future engagement in the platform. Users are more likely to ask another question in the future when the chat room is opened. Third, chat rooms disproportionately benefit new users. When questions are pushed into the chat rooms, askers with lower experience have a higher improvement in question outcome. This paper highlights the importance of a second channel for communication: it complements the main channel in online communities to enhance user engagement and new user inclusion.

Price Obfuscation and Demand Shift: Evidence From a Field Experiment on a Hotel Booking Platform

with Anuj Kummar, Xitong Li and Xiang Wan

Online intermediary platforms are the marketplaces where consumers come to search and compare products from different sellers and make purchases. On one hand, online intermediary platforms may want to facilitate consumers’ price comparisons by listing price information of products from different sellers on the search result pages. On the other hand, online intermediary platforms may want to induce consumers to buy higher priced products by obfuscating prices. Given the trade-off between listing and obfuscating prices, it is important to understand the impact of search cost in price information on sellers and consumers on the online intermediary platforms. To answer this question, we conducted a large-scale field experiment on a leading hotel booking platform in China, examining the impact of hiding price information on the search result pages. Our market-level analysis shows that although hiding price information hurts overall demand, the reduction in demand for hotels with smaller past sales is much smaller than that for hotels with larger past sales. We also conduct user-level analysis and show that treated consumers decrease their booking probability, but the reduction in booking probability is largely driven by a decrease in demand for hotels with larger past sales.