When to Approach the User

Future voice assistants in automated cars are expected to provide a more interactive user experience by making proactive suggestions. For the design of proactive behavior, the right timing of the interaction is key to ensure user acceptance. Context situations during an automated ride are influenced by different factors, which need to be considered to initiate a proactive interaction. The goal of this study is to investigate how different context factors influence the suitability of proactive suggestions and relate to the suggested content. A conjoint analysis was designed to measure users' preferences for different situations. Results indicate that selected context factors are considered suitable or unsuitable for a proactive approach, independent of the suggested content. The content's relevancy contributes to the general acceptance, with more relevant suggestions considered suitable in more context situations. The presented results will inform the modelling approach of a voice assistant prototype for future studies.