The objective of the paper is to analyze the data of existing customers, who bought security solutions in the pension area, to find similarities and purchasing behaviors. The further goal is to provide a decision tree that will tell the customer consultants which clients to target and which product to offer.

The method used to reach the objective is to analyze all the factors influencing the decision process when buying security solutions in the pension area. The initiator provided the necessary data that is analyzed with the help of the software “R”. The dataset for testing will be provided in-house and cannot be used outside the bank for confidentiality reasons.

The goal to match potential clients within their characteristics to a specific pension solution could not be met. Due to the discovery, that around 80 % of customers bought the same product, the recommendation from the data would have been to only sell this product. However, suggestions were made on how this goal might be reached in the future. Firstly, it is suggested to extend the data collection by the factors income, household income, household fortune, education, job title, the indication if the job is related to finance and the declaration if the customer has children. Secondly, it is recommended to save the data in an accessible way in which it can be quickly evaluated afterwards. It was discovered that 81 % of the sales are attributed to Product A. The gender distribution of all sales and per product is almost split equally. For singles, a slight difference between male and female clients is visible in buying interest as well as in risk preferences. The evaluation of the age at which the clients signed up for a pension solution yielded a bimodal distribution peaking at the client’s mid-thirties and mid-fifties.