This dissertation studies new product introductions, in particular the process by which a product is launched and the product exit decision. In the methodological part of this study, a Bayesian learning model with dynamic advertising, pricing, and product exit is developed. The empirical model is intended to explain why firms launch new products, and whether there is evidence that product introductions are market experiments by which firms try to distinguish between profitable and unprofitable products. In marketing, the framework can be used normatively as a decision support tool for managers engaged in the launch of a new brand. The methods developed are applied to the case of the U.S. ready-to-eat breakfast cereal industry, in which product entry and exit is common.
The first part of the dissertation describes the breakfast cereal industry in detail, and discusses the marketing activities of the cereal firms related to brand introductions. The findings from the breakfast cereal industry are compared to other food industries. The second part discusses recent advances in demand estimation, and their usefulness to provide sales forecasts for new products. The last part of this study develops the Bayesian learning model. In this framework, firms may introduce a new product to pursue new profit opportunities, but may initially be uncertain about the profitability of the new brand. If a firm learns from realized sales that the product is a flop, the product exits from the market. The model also allows for an alternative reason to introduce a new product, which is variety-seeking, a taste of consumers for novelty. Using the structural framework developed one can quantify the importance of each hypothesis. The empirical model incorporates a dynamic demand side through long-lived effects of advertising, which is shown to be necessary in order to explain the large volume of advertising spending in the breakfast cereal industry.
My Research