Funded Phd Research
Modeling and Forecasting Buyer and Seller Activity for Two-Sided Businesses
Marketing Department; Faculty Adviser: Peter Fader
There is a large body of research focused upon forecasting future customer activity in traditional one-sided businesses—businesses selling products and services directly to buyers (e.g., bricks-and-mortar retailers). In contrast, there is no well-validated empirical forecasting model for activity in two-sided businesses, for which the firm acts as a platform for both buyers and sellers of goods and services. This is important, because platform businesses (e.g., Amazon, Uber, AirBnB, Etsy, “sharing economy” businesses, and more) are more popular than ever before. Even traditional retailers, such as JC Penney and Nike, are making platforms a key aspect of their strategic vision in coming years.
Our aim is to fill this gap, building an empirical model for how buyers and sellers behave on two-sided markets. We establish the predictive validity of this model by applying our model to a unique large-scale transaction-level dataset from a large retail e-commerce platform business. We then study the substantive implications arising from this model.