Personalizing Consumer Interventions Through Leveraging Behavioral Economics

Abstract

The success of a product is largely based on its customers’ perceived value, which itself is heavily influenced by their beliefs and behavioral biases during the product's lifecycle. Traditional marketing analytics solutions seldom account for such biases and hence yield only suboptimal results. While the field of behavioral economics (BE) has much to offer in identifying human biases in decision making, bias models tailored to specific business problems are hard to come by. Understanding and shaping customer behavior for promoting sustainable products is even more challenging because research on consumer valuation of such products is still in its infancy. Even when interventions recommended by a BE model can be effective at the level of a population, it may not provide desired results at the individual customer level. In this work, we frame our optimal business intervention problem as a reinforcement learning problem, leveraging BE to design the action-space which consists of intervention types that account for consumer biases. Our solution is driven by a use case that aims to promote a new energy storage product based on repurposed EV battery cells. Empirically, we show not only can we achieve substantially better (55%) business outcomes by optimizing our intervention policies, but we can further improve (extra 13%) these outcomes by personalizing these policies using additional customer demographic features.