Social tags are user-defined keywords associated with online content that reflect consumers’ perceptions of various objects, including products and brands. This research presents a new approach for harvesting rich, qualitative information on brands from user-generated social tags. The authors first compare their proposed approach with conventional techniques such as brand concept maps and text mining. They highlight the added value of their approach that results from the unconstrained, open-ended, and synoptic nature of consumer-generated content contained within social tags. The authors then apply existing text-mining and data-reduction methods to analyze disaggregate-level social tagging data for marketing research and demonstrate how marketers can utilize the information in social tags by extracting key representative topics, monitoring common dynamic trends, and understanding heterogeneous perceptions of a brand.