User design toolkits (UDT) and configurators in general are current, well known approaches to satisfying customers’ desires for individualized products. However, the large number of choices and options frequently overwhelm the users and cause frustration and information overload. This phenomenon is known in research as mass confusion (MCF) or product variety paradox and it impedes the effective usage of UDTs. Different papers have identified aspects that decrease MCF, but effective methods for their implementation remain unclear. Recommender systems (RS) are a promising type of technology from the related research area of decision support systems that was proven useful for reducing MCF. Nevertheless, they have not been used in the context of UDTs before. This thesis therefore aims to answer the question of whether or not RS’ are an effective way to achieve focused and need-based navigation, by analyzing the validity of four self- defined hypotheses. The current research about MC and different types of RS was analyzed and used to create a recommender based configurator (RBC) for cupboards and shelves. Afterwards, the prototype was tested in two experiments with groups of students through a comparison with another UDT of the same purpose. The evaluation was built on a path analysis of the collected log data and a small survey. Although the results varied, they show that RBCs can be beneficial for special functions. For example, they can be used as a starting point for configuration and they are helpful for complex parts of products. The users had strong desires to directly influence the created model, which was hard to accomplish by pure recommendations. This thesis indicates that RBCs are an interesting approach to realize an easier, focused and need-based navigation, but are not useful as a complete alternative to the traditional configuration approach. However, they offer great potential as extension.