Comparison of statistical methods for evaluating the effectiveness of marketing campaigns

Marketing is an integral part of everyday life, and its effectiveness is often assessed using various quantitative methods. However, a lack of understanding of basic statistical principles can lead to poor test design, misinterpretation of results, and biased conclusions. This thesis first provides an overview of selected statistical methods suitable for evaluating the success of marketing campaigns and key performance indicators. These methods are then applied to the analysis of a single campaign within Meta Business Manager, followed by a comparison of two campaigns using appropriate statistical tests. Emphasis is placed on designing a methodologically sound approach to hypothesis testing in the context of A/B testing and on selecting appropriate correlation coefficients. Using simulated datasets, the study demonstrates why parametric methods cannot always be applied and highlights the benefits of non-parametric approaches. Finally, the thesis proposes effective procedures for evaluating and visualizing the results of the analysis.