| The ubiquity of the Internet and the pervasiveness of Web 2.0 technologies have facilitated the emergence of user generated content in various forms and in multiple outlets. Users contribute their knowledge and expertise, share resources and ideas, and evaluate online content. Recognizing the profound impact, firms have been increasingly leveraging it. They go beyond merely hosting this content and start actively soliciting it. In this dissertation, I study two contexts of user generated content hosted by firms: technical support forums and opinion forums. In the first essay, I study technical support forums where IT vendors host a medium for users to solve each other's problems. Firms influence user behavior by implementing incentive mechanisms that promote the generation of solutions. Using datasets collected from four major forums, I derive the role of such incentive mechanisms on the quality and effectiveness of user generated content. I also analyze the impact of these mechanisms on the overall forum quality and on the effectiveness of technical support. I show that firms implementing quality based incentive mechanisms promote not just larger quantity of user contribution but also higher quality ones. In the second essay, I analyze the delay incurred by users to obtain a solution in technical support forums and devise a policy for firm involvement to reduce this delay. Using a combination of empirical analysis and queueing theory, I find that users who post questions suffer two types of delay, one type incurred by initiators (those users who initiate threads) and another incurred by joiners (those who join midway in the life of a thread). I find that joiners wait more than initiators. I address this counter-intuitive result by formulating a firm involvement policy that reduces joiners delay. In the third essay, I analyze opinion forums in a unified view that incorporates both word-of-mouth (WOM) and system recommendation. In particular, I focus on the competition effect arising from these two tools. I account for various endogeneity effects emanating from omitted variables and simultaneity which have been mostly overlooked in the literature. I derive the cannibalizing effects that recommender systems have on the performance of book sales in a competitive market. I also highlight the role of opinion leaders in contributing to better sales. |
