Abstract Elasticity estpdeltimates provide the brand manager with useful diagnostics for evaluating competitive market structure. However, an econometric model must often be simplified due to the limited amount of data available to estimate the model's parameters. This results in a reduction in the structural insight one can gain from the model. Capitalizing on the forecasting ability of neural networks, we introduce an innovative method of extracting elasticity structure for a convenient consumer retail product market. The resulting forecasting measures and elasticity structures are then compared with those obtained from a differential-effects multiplicative competitive interaction (MCI) aggregate market share model. We find that the neural network slightly outperformed the differential-effects MCI model with regards to model fit. Our results also suggest that the neural network offered superior estimates of asymmetric cross-elasticities which resulted in superior forecasting ability of the holdout sample.