Essays on the role of social interactions and networks in marketing

This dissertation explores the role of social interactions and networks in three important and relatively unexplored spheres in marketing: fashion, blogs and online video sharing. In the first essay "Cloak or Flaunt---The Firm's Fashion Dilemma", we explore why fashion firms often strategically withhold information on the 'fashion hits'. We model fashion as a facilitator of social interactions wherein (1) it allows consumers to signal to each other that they are 'in the know', and (2) it enables them to fit in with their peers. We model how the firm's communication interacts with consumers' value of fashion as a social device and derive the conditions under which less information can be more profitable from the perspective of a firm marketing conspicuously consumed goods.;The second essay "Link to Success---How Blogs Build an Audience by Recommending Rivals" explores the rise in popularity of blogs. It explains why bloggers often recommend rival blogs at the risk of losing readers to them. In our model, blogs differ along two dimensions---(1) ability to produce information, (2) ability to link to informative blogs. While links demonstrate a blog's ability to link, it also reveals the information producing ability of a competitor. In equilibrium, blogs that reliably produce information have more incoming links and visitors. Thus, this essay studies the trade-offs inherent in recommending rivals. We also empirically test the model's predictions.;The third essay "Impact of Social Network Structure on Content Propagation: A Study Using YouTube Data" explores how social networks affect popularity of YouTube videos. We present a dynamic model of video growth using data on a panel of videos and the social network of the videos' authors. We derive metrics to quantify three aspects of network structure: (1) Connectivity, (2) Clustering, (3) Centrality. We find that network structure plays a significant role in video growth. Specifically, connectivity has a positive impact, while clustering and centrality have a negative impact. Also, network properties play very different roles in early vs. later stages of growth. Our results demonstrate the power of social networks in driving video popularity.