For predictive analysis, whether simple or highly sophisticated, a model of the data is needed for future projections.   Our experience is that the power of predictive analysis is becoming more refined and easier to validate assumptions as more data is included in the process.   We apply simple models to fit data for some applications and more complex agent-based simulations for others.   The value of big data is that many scenarios can be examined and updated as data becomes available in real or near-real time.   Mission applications to date include planning, simulation of outcomes and impacts of decisions affecting individuals and population, and scenario developments of market resiliency and community acceptance.