In this paper, I develop afresh a comprehensive approach to market segmentation of Facebook users. The approach considers the implications of Facebook's business model built around its multi-sided platform (MSP). In the first part of the study, I present a series of five propositions that inform the formulation of segmentation and targeting strategies for Facebook and more generally MSPs. These are based on a review and synthesis of extant literature on multi-sided markets and classification of Facebook and social media users. The second part of the study uses empirical survey data from 261 Facebook users to (a) evaluate some of these propositions, and (b) develop a segmentation scheme that may guide the development of marketing strategy for social networking sites (SNS) like Facebook. The empirical analysis makes use of cross-tabs, classification and regression trees, linear discriminant analysis, cluster analysis, and artificial neural networks (ANN). The basis variables of importance to Facebook for market segmentation,according to the findings, are frequency of use and level of engagement. Important descriptors that help determine segment membership are: marital status, blogging habits, use of Facebook mobile, general interest in online social networking (OSN), and use of competing OSNs such as LinkedIn and Google. I conclude with a discussion on future research priorities in strategizing for social networking sites.