Market research is vital for company business good development, for providing goods delivery on internal and external market, for micro and macroeconomic substantiation and justification of investments, for long run profitability microeconomic programming. Statistical instruments are used in several phases of market research, since the phase of working out the research program till the phase of interpretation and analysis of market information, supplying company's manager the possibility of the rightly decisions-making. The correlation and regression techniques aim at defining the relationship notion between a dependent variable and one or several independent variables. Marketing forecasts represent the estimations of the future marketing variables. In this way, marketing forecasts are attempts to forecast the future of marketing variables, on the basis of their past estimation. Marketing decisions quality concerning the economic activity management depends upon the forecasting accuracy. Marketing forecasts can be performed at the level of product, group of products, economic unit, activity branch, or of certain territorial-administrative units, national, international level etc. Forecasts are difficult to be achieved due to demand instability, but they are important to company success. Forecasting can be made in three phases: macroeconomic forecasting; industrial forecasting; company sales forecasting. Macroeconomic forecasting refers to inflation, unemployment, interest rate, consumption expenses, exports, imports etc., the final result being the GNP. On the basis of GNP, forecasting is made at industry level, the company going to deduce its own sales, according to market shares. Forecasting uses three types of information: consumers' opinions, known by means of surveys on buyers' intentions, by selling agents' opinions, by experts' opinions; consumers' behaviour settled by a market test, in order to see which are the buyers' reactions; consumers' previous behaviour involving the analysis of previously buying behaviour. This can be analyzed by using the multiple regression that considers the future and the past sales as a time function, finding the factors which affect the sales and their relative importance.