Speaker: Manish Raghavan
Location: The seminar will take place in the usual room, SEC 1.413, and will also be streamed on zoom here. Hope to see you there :).
Title: The Challenge of Understanding What Users Want: Inconsistent Preferences and Engagement Optimization
Abstract: Online platforms routinely optimize the content they recommend to their users based on behavioral data. At its core, the use of behavioral data is predicated on a revealed preference assumption: users choose what they like; and thus what they choose reveals what they like. And yet, research has repeatedly demonstrated that behavior can be a poor proxy for users’ preferences, exactly because users themselves are conflicted. Indeed, our own intuition tells us that we often make choices in the moment that are inconsistent with our long-term preferences. The behavioral economics and psychology literatures suggest that this is because we don’t have a single set of preferences that governs our behavior; instead, our choices are based on conflicting desires.
In this work, we develop a model of media consumption where users have inconsistent preferences. We consider an altruistic platform who simply wants to maximize user utility, but only observes behavioral data in the form of the user’s engagement. We show how our model of users’ preference inconsistencies produces phenomena that are familiar from everyday experience but difficult to capture in traditional user interaction models, These phenomena include users who have long sessions on a platform but derive very little utility from it; and platform changes that steadily raise user engagement before abruptly causing users to go “cold turkey” and quit. A key ingredient in our model is a formulation for how platforms determine what to show users: they optimize over a large set of potential content (the content manifold) parametrized by underlying features of the content. Whether improving engagement improves user welfare depends on the direction of movement in the content manifold: for certain directions of change increasing engagement makes users less happy, while in other directions on the same manifold increasing engagement makes users happier. We provide a characterization of the structure of content manifolds for which increasing engagement fails to increase user utility. By linking these effects to abstractions of platform design choices, our model thus creates a theoretical framework and vocabulary in which to explore interactions between design, behavioral science, and social media.