The forecasting accuracy of market share models using predicted values of competitive marketing behavior

Abstract The predictive power of market share models is a subject that has received considerable attention in marketing literature. However, hardly any attention has been paid to the question of how the values of the marketing instruments of competitors can be predicted. This is remarkable since these values constitute the input variables for a market share model. In this paper we investigate the sensitivity of predicted market shares to different assumptions with respect to competitive behavior. Calibrating a number of models and predicting competitive behavior and market shares for nine brands from three markets using bimonthly data leads to four conclusions: (1) market share predictions are not very sensitive to the assumptions that are made with respect to future competitive marketing behavior; (2) sophisticated predictions of competitive behavior are not consistently more accurate than naive predictions; (3) using predicted values of competitive behavior may provide better market share predictions than observed values of competitive behavior; and (4) sophisticated predictions of market shares are not systematically better than naive market share predictions.