Speakers in research dissemination talks are challenged with the need to connect with an audience that does not necessarily share their knowledge and expertise. This communicative situation can be particularly challenging for speakers using English both as a foreign language and for academic purposes. This study combines multimodal and ethnographic methods to explore how speakers of dissemination talks engage with their public. It focuses on four presenters’ use and combination of language, paralanguage, kinesics, proxemics and gaze during intensive moments of engagement. The results show that these interpersonal rich points consist of dense multimodal ensembles that serve to shorten the distance between presenters and their audiences. The findings suggest that a skilful orchestration of modes can be greatly beneficial to achieve the desired level of audience engagement. Therefore, developing speakers’ multimodal communicative competence should be a priority in English for specific and academic purposes (ESP/EAP) training.