Forecasting the demand for new telecommunication services

This paper proposes a novel forecasting method which is applicable for new services, where little historical data has been recorded. The method uses instead estimators based on economical, demographic and traffic data. The method is, compared to traditional forecasting procedures that are built upon a solid historical record of data, clearly found to be weaker numerically. However, for novel services it obviously has the advantage that it provides a result, where the traditional method would fail because the data record is too weak. Furthermore, our method has the advantage that as soon as further data becomes available, it can be refined to provide even better results.