Mathematical Modeling of Competition in Sponsored Search Market (1006.1019v2)

Abstract: Sponsored search mechanisms have drawn much attention from both academic community and industry in recent years since the seminal papers of [13] and [14]. However, most of the existing literature concentrates on the mechanism design and analysis within the scope of only one search engine in the market. In this paper we propose a mathematical framework for modeling the interaction of publishers, advertisers and end users in a competitive market. We first consider the monopoly market model and provide optimal solutions for both ex ante and ex post cases, which represents the long-term and short-term revenues of search engines respectively. We then analyze the strategic behaviors of end users and advertisers under duopoly and prove the existence of equilibrium for both search engines to co-exist from ex-post perspective. To show the more general ex ante results, we carry out extensive simulations under different parameter settings. Our analysis and observation in this work can provide useful insight in regulating the sponsored search market and protecting the interests of advertisers and end users.