Two essays on measuring and managing the influence of blogs

This thesis investigates how firms can measure and manage influence of IT-enabled communication media, particularly blogs. The thesis constitutes two essays. The essays can be read independently, but they are complementary to each other. ^ The first essay measures the influence of employees\u27 negative blogging on their employers and suggests an approach to govern employees\u27 negative blogging for better results. Results indicate that the increase in negative posts, increases readership exponentially up to a certain level of negative posting, and then stabilizes beyond this point. Due to this initial exponential increase in readership, there are conditions under which negative posts generate greater net positive influence on readers than if there were no negative posts. We illustrate the application of the framework using blogging data from employees at Sun Microsystems. Our empirical model accounts for inherent non-linearities, issues of endogeneity and unobserved heterogeneity, and potential alternative specifications. ^ The second essay investigates the influence of blogs in general new venture financing and proposes an approach to handle blogs for better financing deals. Findings indicate that eWOM of popular bloggers helps new ventures in getting higher amounts and valuations, and eWOM due to non-popular bloggers does not matter. Other findings are that the impact of negative eWOM is more than the positive eWOM. Results also show that as new ventures progress through financing stages, the effect of eWOM on financing decreases. Another contribution of this work is to gather and assimilate data from wide variety of sources such as VentureXpert, surveys, Google Blogsearch, Lexis-Nexis, Archive.org, and Hoover. Econometric analysis is designed to control for biases in sample selection, simultaneity in financing variables, and potential alternative specifications. Robustness checks such as ignoring selection bias fix, controlling for potential endogeneity bias due to unobserved quality, adding more control variables, and by dropping ventures that received multiple rounds of financing were conducted.