Dynamic Competition between Data-driven Platforms, JMP (draft coming soon)
In this paper, I develop a dynamic model where two data-driven platforms compete for users' attention. The quality of the service on each platform improves as more user data is available. At each period, platforms choose between current monetization through advertisement, or the accumulation of more user and therefore more data, in the future. Accordingly, I relate the market outcomes in equilibrium to platforms' advantages of initial market share and capability to process data. I show market co-participation occurs when data productivities of both platforms are small. Market tipping is more likely to occur when one of the platforms has a large data productivity and enjoys a large user base advantage. I also explore the effects of compulsory data sharing. Market tipping is less likely to occur with data sharing. Besides, data sharing by both platforms, or by the platform with a large technology advantage increases consumers' surplus.
Complementary Multi-Sided Platforms, with Doh-Shin Jeon, Yassine Lefouili and Timothy Simcoe (draft coming soon)
Motivated by several examples, including Internet of Things (IoT) patent licensing, we consider a model in which m (≥ 1) complementary platforms choose prices for n (>1) connected devices which generate demand externalities among themselves. We characterize equilibrium prices and show that platforms face a trade-off between an externality internalization effect and a value extraction effect. Given a device, the externality internalization effect (the value extraction effect) represents a weighted sum of all the externalities that a device generates to (receives from) other devices. The weight assigned to each device reflects its position in the network of demand externalities and is measured by the Katz-Bonacich centrality. We show how the centrality measures and resulting pricing decisions change with the number of platforms. Even if Cournot’s insight continues to hold (complementary monopolists charge higher prices than an integrated monopolist), surprisingly, the total prices for some particular devices in a duopoly can be lower than the prices in the single monopoly benchmark. We contribute to the two-sided market literature by analyzing complementary platforms in a general multi-sided market.