Business and Consumer surveys are designed to signal turning points and provide in advance the informatik about potential changes in the economic cycle. The authors, using advanced methods of time series analysis, especially Granger causality and vector autoregressive models, deal with the question of to what extent the results of the surveys in the form of confidence indicators are able to outpace the development of the Czech economy represented by gross value added. In addition, the authors, experimenting with the structure of surveyed questions and used weights, propose some modifications in the construction of confidence indicators as stipulated by the European Commission with the aim to improve their forecasting abilities.