This paper suggests that humans value robotic systems as signals—costly, visible commitments that can secure access to preferred resources. This contrasts with HRI research, design, engineering and deployment, which have focused on robots’ instrumental value—namely how designing and interacting with them may produce more or less productivity. Drawing on a multiyear ethnography of a “failed” robotic telepresence deployment in a teaching hospital, this paper shows that robots’ signaling value can significantly outweigh—and even contravene any practical utility that they may provide through use. This analysis further suggests that unlike nontechnological organizational signals robots’ signaling value is highly contingent on the observability of their use. Considering robots’ signaling value in complex social systems such as organizations promises improved robotic systems development and deployment techniques and improved prediction regarding human-robot interaction.