Data Mining becomes a vital aspect in data analysis and clustering is a potential tool of Data Mining. In this work we apply the Data Mining method of an Artificial Neural Net, namely Self-Organizing Maps, on a dataset containing information about new business contracts. We explain neural nets in general and how Data Mining can be applied in insurance companies. By means of the classical vector quantisation process we explain the algorithm of Self-Organizing Maps and its parameters. We then apply the algorithm to a data sheet provided by the actuarial life department of Allianz Elementar Lebensversicherungs-AG to get a better insight into parameters affecting the new business margin, going beyond the already widely performed analyses. The outcomes show clear evidence that Self-Organizing Maps can cluster this data into individual groups. Though the results support the assumption of a highly non-linear correlation between the new business margin and the parameters leading to it, only a small amount of data can be used for analyses. A suggestion on how to set parameters leading to a more stable, higher new business margin is made nevertheless. Because of limitations concerning data, software and time, Self-Organizing Maps are not the ideal solution for analysing this kind of data. Especially due to the required time-consuming, manual pre- and postprocessing it is not recommended to use the methods presented in this work on a regular basis on this particular data sheet.