What is currently unknown, is the level of knowledge and demand of the Baloise user base regarding automated communication channels and chatbots. If the demand level is not sufficient or the user requirements are not fulfilled, the service desk might invest in a chatbot solution without benefitting from the strengths, which a chatbot has to offer. The author of this thesis is to solve this problem, by assessing the knowledge and demand level of the user base regarding the topic chatbots. The author also faces the task of gathering the requirements for a chatbot from a user's point of view.

At first, the reader is presented with an introduction into the topic of chatbots. Going further, two sets of interviews are held, and a survey is conducted. The survey and the first interview set are done to assess the current level of knowledge and demand of the user base. Afterwards, the requirements for the chatbot from the service desk project are collected and a chatbot prototype is developed. The list of requirements is then extended with the users’ requirements after a presentation of the prototype. The refined list of requirements is then prioritized according to the Kano model.

The result of this bachelor thesis is an analysis of the Baloise user base regarding the topic of chatbots. The analysis is based on a balanced number of quantitative and qualitative information gathering methods. The requirements from the viewpoint of the service desk project as well as the user base are collected, structured and compared. Supported by the previously gathered data, a recommendation is given to the service desk. The recommendation includes user knowledge and demand levels, ethical concerns, reactions to a chatbot prototype, and prioritized requirements. This recommendation is used by the service desk to adapt the chatbot project to fit the users wishes before the go-live of the solution later this year.