.

In my dissertation I am interested in the management of marketing cummunication and techniques which make this communication more effective. The main topic of this thesis is the optimization of marketing communication which is discussed from the theoretical point of view at first. Then a social network analysis improving additional and complementary information to predictive models related to customers' behaviour is a part of this thesis. The analysis of social network represents an approach which is used for study of social structure and the analysis of social linkage of individuals in a form of relational data is a subject of its interest. Models of binary choice are used to estimate predictive models, the main attention is dedicated to the logistic regression, further to the logit and probit models including discussion related to the possible use of models of multinomial choice. The mentioned techniques are combined in the application part. The telecommunication market is selected for case studies because it represents a dynamically developing market opened to new and fresh technologies and approaches. Three case studies are discribed in this thesis. The first study is focused on the optimization of proactive telemarketing campaigns realized with foreign telecom operator. The second case study is focused on the optimization of reactive telemarketing campaings what means the optimization of marketing communication in call centers in czech telecom operator. The third case study relates to the important change of czech telecom market which happened in April 2013 and to the role of predictive modelling and optimization in it.