The Dynamic Effect of Innovation on Market Structure

Abstract

Product innovation is endemic among consumer packaged goods firms, and is an integral component of their marketing strategy. As innovations impact markets, there is a pressing need to develop market response models that can adapt to such changes. Our model copes with the challenges dynamic environments entail: nonstationarity, changes in parameters over time, missing data, and cross-sectional heterogeneity. We use this approach to model sales response in the frozen pizza category, in which the introduction of rising crust pizza brands represented a major innovation. The model is directly applicable to other domains in which market structure might be non-stationary: changes in promotion strategy, shifts in the retail environment, movements in macro-economic factors, etc.

In our application, we find that innovation i) makes the existing brands appear more similar, as indicated by increasing cross-brand price elasticities, ii) decreases brand differentiation for the existing brands as indicated by an increase in the magnitude of own-brand price elasticities and iii) increases the variance of the sales response equations temporarily around the time of the introduction of the innovation, indicating increased uncertainty in sales response. The adjustment of the market to the innovations takes about seven weeks.

We conclude by discussing the managerial implications by i) presenting maps of how clout and vulnerability evolve over time, ii) assessing the effect of new brands on cannibalization and iii) considering the strategic implications of introducing a flanker innovation to facilitate the ability of an extant brand to attack an extant incumbent leader.

Keywords: Innovation, Brand Entry, Market Structure, Frozen Pizza, Time Series Methods, Dynamic Linear Models, Bayesian Methods, Gibbs Sampler, Long-term Effects