This research essay highlights the need to integrate predictive analytics into information systems (IS) research. Predictive analytics include empirical methods that generate data predictions and assess predictive power. They are essential for creating practical models and complementing explanatory modeling in theory building and testing. The paper describes six roles for predictive analytics: new theory generation, measurement development, comparison of competing theories, improvement of existing models, relevance assessment, and assessment of the predictability of empirical phenomena. Despite their importance, predictive analytics are rare in the empirical IS literature, which relies on explanatory statistical modeling. Explanatory power does not imply predictive power, so predictive analytics are necessary for assessing predictive power and building empirical models that predict well. The paper discusses differences between predictive analytics and explanatory statistical modeling, showing how they lead to different final models. It also presents methods for assessing predictive power and building predictive models, illustrated by converting a well-known explanatory study of TAM into a predictive context. The paper concludes that predictive analytics add theoretical and practical value to IS research.This research essay highlights the need to integrate predictive analytics into information systems (IS) research. Predictive analytics include empirical methods that generate data predictions and assess predictive power. They are essential for creating practical models and complementing explanatory modeling in theory building and testing. The paper describes six roles for predictive analytics: new theory generation, measurement development, comparison of competing theories, improvement of existing models, relevance assessment, and assessment of the predictability of empirical phenomena. Despite their importance, predictive analytics are rare in the empirical IS literature, which relies on explanatory statistical modeling. Explanatory power does not imply predictive power, so predictive analytics are necessary for assessing predictive power and building empirical models that predict well. The paper discusses differences between predictive analytics and explanatory statistical modeling, showing how they lead to different final models. It also presents methods for assessing predictive power and building predictive models, illustrated by converting a well-known explanatory study of TAM into a predictive context. The paper concludes that predictive analytics add theoretical and practical value to IS research.