Point-of-sale scanner data provides a unique opportunity for analyzing consumer package goods (CPG) trends and patterns. Decisions of ever-increasing complexity are made possible by the amount of data available. Unfortunately, the analysis of such voluminous data requires complex techniques and processing requirements not available to many marketing decision makers. In this paper, we describe a prototype system which allows users to manage the complex models and scanner data to make forecasts in an interactive fashion. A limited test of the prototype allowed users not familiar with the underlying models to develop product forecasts.