Abstract
Online marketing is characterized by a loss of control when compared to traditional offline marketing (Mangold & Faulds, 2009). This is mainly due to two reasons. The first is that, on online social networks, communities took control over brands. The second is that online social network (OSN) platforms, mainly Facebook, limit access to information in a way they do not even have access to their list of fans.This dissertation comes at a time where more research and more investigations are needed in the field of strategic marketing on online social networks (Hoffman & Novak, 2012). This is why it attempts to add to theory and practice of viral marketing strategies on OSN and, in particular, initial seeding population. It is an attempt to reestablish equilibrium to the advantage of practitioners in marketing on OSN and increase their margin of manoeuver between a platform that builds barriers to information and a public who considers brands on OSN as intruders of its own space (Fournier & Avery, 2011). We chose the critical approach of Habermas (1985). In this paradigm, knowledge is based on empirical experiments where environment tweaking is allowed.The analysis and experiments in this dissertation follow a theoretical and experimental path that leads to a theoretical and practical thinking that defines strategies and viral marketing optimization methods through the control of viral marketing elements, especially the selection and creation of initial seeding populations.This dissertation consists of several chapters, amongst which a paper currently under third revision at the Journal of Advertising Research (JAR) and four experiments that have been presented at international conferences (Euras, 2012; INSNA, 2014; ISMS, 2015) and the workshop stratégie Paris-Dauphine (2014).This dissertation attempts to answer several questions about the control of viral marketing elements, especially the optimization of initial seeding population, and we think we have found relevant and innovation answers to these questions. We start with a meditation on research on online social networks and, in particular, the ethics of research on the internet and OSN. We then move to the question of uncovering and building social graphs of fans and analyze the efficiency and inefficiency of fake profiles in Facebook marketing. We also consider word-of-mouth in the context of an optimized seeding population, and the discovery of social bridges over structural holes rallying antagonistic political groups, a theoretically optimized seeding population. Those experiments lead to a thinking about the importance of the Facebook channel (pages, groups or personal timelines) and the role it plays with different marketing strategies: mass, relational, niche and, especially, viral.The last chapter puts into practice our findings and results by creating a strategy from zero to success of a personal branding campaign using viral marketing. This experiment is document, analyzed and without bias because it only happens on Facebook and does not use any advertising sources inside or outside this channel. This last experiment shows the efficiency of the viral marketing strategy based on our results, finding and theoretical analysis in this dissertation.