Essays on Dynamic Marketing Intercommunications.

Abstract

The dissertation examines two distinct problems related to “marketing communication dynamics”. The main goal of this line of research is to help firms provide individually tailored marketing contents to their customers. In these two essays, I develop statistical models to first understand customers’ responses, and then explore methods to optimize firms’ reactions accordingly. Essay 1 examines “scale attraction effects” in a charitable donation context, introducing novel constructs (“compliance degree”, “pulling amount”, “accumulated pulling amount”) to describe attraction effects for multi-point appeals scales. The proposed model jointly accounts for donation incidence and amount using a Tobit 2 formulation, and allows heterogeneity in seasonality and pulling effects. Results suggest substantial scale attraction effects that vary across donors, stronger “pulling down” than “pulling up”, and heterogeneous seasonal donation patterns. A significantly negative error correlation between donation incidence and donation amount underscores the importance of accounting for selectivity effects. The effects of individually tailoring appeals scales is demonstrated through simulation. Essay 2 investigates mate-seeking users’ decision rules in an online dating context. I develop an empirical two-stage mate choice model that can accommodate compensatory and non-compensatory decision rules in each of two stages: browsing and writing. A mixture of logits model with changepoints allows for distinct decision rules across stages and heterogeneity in rule use across site users. Most importantly, it allows us to identify and compare attribute-level decision rules (“deal-breakers” and “deal-makers”) over the two stages. Results suggest the existence of heterogeneity in decision rules across (1) genders, (2) stages, and (3) site users. Additionally, it suggests the existence of potential deal-breakers/makers across both discrete and continuous attributes.