In order to arrive at statistical forecasts of the future growth of markets it is necessary to use statistical and econometric methods which are regarded in many quarters as being worlds apart from the conventional (and often qualitative) methods used in market research. It is difficult to combine and illustrate by quantifiable results, those market trends which are dictated by changes in the general economic climate and those which result from measures; taken by companies themselves. Now, if market research is not to be of limited value, it is necessary to evaluate the combined effects of both trends so as to be able to make an accurate assessment of the risks involved in any decision to take a specific course of action. In this respect a very great deal remains to be done, particularly as far as industrial products are concerned. The conclusions contained in this article are the fruit of several years first-hand experience and are intended: - to show what results can be achieved by the use of the well-known forecasting methods; - to stress the lines along which research should be conducted in order to evaluate companies' development policies.