Improving Match Rates in Dating Markets Through Assortment Optimization

ABSTRACTMotivated by our collaboration with a major US online dating company, we study how a platform should dynamically select the set of potential partners to show to each user in each period to maximize the expected number of matches in a time horizon, considering that a match is formed only after two users like each other, possibly in different periods. Increasing match rates is a prevalent objective among online platforms. We provide insights on how to leverage users? preferences and behavior towards this end. Our proposed algorithm was piloted by our collaborator in major cities in the US. We introduce a model of a dynamic matching market mediated by a platform. The platform hosts a set of users and must decide, in each period, what subset of profiles to show to each user to maximize the overall expected number of matches. Each period, users log in with some time-dependent probability and, conditional on logging in, observe a set of profiles-an assortment-that satisfies the constraints imposed by the platform. Then, users decide whether to like or not like each profile in their assortment based on their preferences. If two users mutually like each other, possibly in different periods, a match is generated. Our goal is to find an algorithm to maximize the total expected number of matches generated by the platform over an entire time horizon. We show that the platform's problem is computationally hard.