Functional Regression: A New Model and Approach for Predicting Market Penetration of New Products

The Bass model has been the standard for predicting the market penetration of new products. Recently a new class of non-parametric techniques, Functional Data Analysis (FDA), has shown impressive results within the statistics community. The authors test the predictive performance of FDA versus variations of the Bass model on 760 new categories across numerous products and countries. Ten fold cross validation shows that Augmented Functional Regression provides distinctly superior predictions by using a) four underlying dimensions of the pattern of penetration across 760 categories, b) deviations of a target category from the mean on these four dimensions, and c) knowledge of the target product.