The emergence of product certifications such as "organic," "fair trade" and "shade grown" has gone far to introduce non-price related product information back into the consumer decision making process thereby reducing information asymmetries and improving market outcomes. Recently, these certifications have been supplemented by the emergence of several web-based consumer decision support systems that are designed to disseminate non-price product information to consumers. Like printed product labels, these new web-based systems present the same informational cues to each consumer regardless of his or her preference for the information and they rate products with fixed weights that are not customized to reflect the values of individuals. This research proposes and evaluates an automatable process by which consumers’ non-price information preferences can be measured thereby allowing a web-based system to provide them with customized information packages that include only information that is relevant to their individual decision processes and is formulated in a way that is consistent with their personal preferences. Second, this research identifies a set of information policies that may provide guidance for the design of such an information system. This research was performed by integrating a judgment analysis aimed at measuring subjects’ preferences with a simulation environment calibrated to a single product, Mexican coffee. Results support the need for non-price product labeling and revealed that, to certain subjects, non-price information was actually more important than retail price. The results also revealed challenges to current product labeling regimes including the need for customized information packages, the need for labels that include information about the welfare of all supply chain entities, and the need to disaggregate certain existing certifications into their core components. Further, the results offer recommendations for the design of online consumer decision support systems including insights into how to manage the stability versus accuracy tradeoff. Finally, a set of methodological findings are presented to guide the design and use of future linked simulation-judgment analysis models.