Activating users on online platforms is a critical endeavor that requires the employment of adequate user onboarding strategies, which focus on converting visitors into revenue-generating users. Despite a robust understanding of the antecedents of user onboarding behavior, researchers have devoted only little attention towards how platforms can actively influence desired user onboarding outcomes. Drawing on social response as well as social exchange theory, this study examines how disembodied interfaces like chatbots can facilitate the user onboarding process. In cooperation with a German startup company, we empirically tested in a randomized field experiment with 2095 visitors how low vs. high message interactivity (i.e., static vs. conversational presentation of requests) and platform self-disclosure (i.e., a platform providing information about itself) affect user disclosure propensity (i.e., likelihood that a user discloses information). Our results demonstrate that users in high message interaction conditions were significantly more likely to self-disclose in contrast to low message interaction conditions, while platform self-disclosure had a significant positive effect as well. Furthermore, high message interactivity significantly amplified the effect of platform self-disclosure on user disclosure propensity in contrast to low message interactivity. Consequently, our study provides novel findings on the effectiveness of disembodied interfaces to improve user onboarding behavior.