This paper discusses the (divergent) impact of written versus spoken feedback (and matchers' number) on communicators' efficiency in a shared (card shape recognition) task across different modalities (presence in situ and via streaming video chat).