Date of Completion

4-3-2020

Embargo Period

4-1-2030

Keywords

Relationship marketing, Structural model, Salesforce management, Social media advertising, Hidden Markov model, Cross effect

Major Advisor

Debanjan Mitra

Co-Major Advisor

Hongju Liu

Associate Advisor

Joseph Pancras

Associate Advisor

William T. Ross Jr.

Field of Study

Business Administration

Degree

Doctor of Philosophy

Open Access

Open Access

Abstract

A cross effect refers to the effect that a marketing strategy developed for a brand or targeting an individual has on other brands or individuals. Without taking into account this unintended “externality”, decision makers may end up with suboptimal outcomes. In this dissertation, we study the cross effects of the marketing mix variables in both business markets and consumer markets. The first essay explores the long-term unintended impacts of relationship management activities on the performance of business partners. We develop a bivariate Tobit hidden Markov model to capture the impact of relationship marketing on the dynamics of relationships between companies and independent sales agents. Calibrating our model with a comprehensive dataset provided by a leading insurance company in the US, we find that the relationship management activities targeting one agent have negative impacts on her colleagues while the sales performance of one agent has positive impacts on other agents’ performance. The second essay examines the cross effects of TV and owned social media (OSM) advertising in the U.S. soft drink industry. Using a random coefficients logit model, we investigate within a brand portfolio, 1) how one brand’s OSM advertising spills over to other brands, 2) how OSM advertising spillover differs from the spillover in TV advertising, and 3) how OSM and TV advertising jointly influence consumer purchase. We find that advertising one brand on TV increases sales of other brands owned by the same company, while one brand’s OSM advertising restrains other brands’ in the same portfolio.

Download

Available for download on Monday, April 01, 2030

COinS