Comparing naive with econometric market share models when competitors' actions are forecast

While econometric market share models have been shown to be useful to managers as descriptive tools, controversy exists over their use in forecasting. As a result, a criterion as been developed which can be used to assess whether or not a naive model will outperform an econometric model for market share forecasting. However, this criterion ignored the effects of having to forecast competitors' actions, as would be demanded in a real-life market share forecasting application. In this paper previous results by the author are extended to cover the situation where competitors' actions must also be forecast. The results show that having to forecast competitors' actions is very demanding for econometric models. Consequently, the naive model is likely to be preferred in most market share forecasting situations where competitors' actions are forecast. Econometric models will be preferred only when they fit the data extremely well for all brands in the market.