| Abstract [eng] |
The main aim of this thesis is to conduct a segmentation of clients and churn detection for a most profitable segment of a printing company. The thesis consists of three main parts: a literature review, investigation of methods used at work and a research part in which the results of a research are presented. A discussion about customer segmentation and loyalty is presented in a literature review part as well as a short review of the most popular customer types and their most distinctive features. In this part there is also an analysation of most loyal customers separating them by distinction of their characteristics, benefits and the opportunities of encouragement loyalty. In the second part of this thesis customer segmentation and loyalty detection methods, such as ABC analysis, RMF analysis, Two Step cluster analysis and machine learning algorithms for detection were carried out using the algorithms of classification. Also, the aims of using above mentioned tools, their operating principles and the examples of calculations which were found in literature are defined in the second part. In the third part, the overview of clients and the outcome of implementing the methods mentioned in second part is presented. Several customer segmentation cases are carried out dividing them in to various clusters using variables of RFMT. During next step customer loyalty study is conducted to the most valuable clients from segments A and B, which were determined after carrying on ABC analysis. Using machine learning algorithms together with cross validation the loyalty of clients is evaluated by solving detection tasks. After conducting analysis of clients, the recommendations are suggested to the company. |
