As marketers manage more products and channels, it’s not enough to model top-selling items in supermarkets. Yet there are barriers to expanding coverage, including overly aggregate data, missing causal, and slow movers. Potential solutions are additive models, grouping data to maximize homogeneity, pooling or Bayesian techniques for sparseness, and imputed causal. For Wal-Mart, the best strategy is to model Retail Link® data. We believe marketers will find the benefits of broader coverage justify efforts to tackle data and model challenges.