The performance of a microbrewery is heavily dependent on its location. Badly chosen locations, often due to ad-hoc or poorly conducted site selection analysis, are a key factor for businesses to go down. The emergence and rapid development of Geographic Information Systems (GIS) with their spatial analytical capabilities open new doors to strategic site performance analysis. Combining GIS analysis and mathematical models with geodemographic and social media data can help planners and decision makers in the essential evaluation process of potential candidate sites. The focus of this research project lies on gaining new knowledge about craft beer consumption characteristics and behaviors, and utilizing this knowledge to develop of a custom site selection model for microbreweries which calculates the number of potential craft beer consumers within a specified geographic area. The study area comprises the entire County of San Diego. An exploratory analysis is conducted in order to gain new knowledge of craft beer consumers based on a set of Twitter data collected over a time period of one year. All tweets contain locational, temporal and textual data on which spatial, qualitative and quantitative analyses methods are applied to obtain information about the spatial distribution and demographic characteristics of craft beer consumers. The demographic information gained from the exploratory analysis is used for the development of the site selection model which corresponds to a geodemographic probability model taking as an input the demographic characteristics of a specific geographical area. The model is applied in two location analysis scenarios at different scales. The first one is a macro level analysis with the objective of estimating the number of potential craft beer consumers for each ZIP code in San Diego County. The second scenario corresponds to a micro level site suitability analysis, evaluating the feasibility for opening a new microbrewery on a certain number of vacant sites which are either for sale or rent.