When designing or redesigning socio-technical systems, it is often required that those systems be more ‘resilient’ as a result. However, exactly what is meant by resilience in these contexts is unclear. To design resilient systems, we must first be able to answer a number of questions, including: Should a resilient system change to accommodate influences or stay the same? If the system changes, where should this change take place? How do we decide which system, or sub-system, to make resilient? For any given system, answering these questions requires engagement with different stakeholders, allowing a conversation to take place that typically spans different disciplines. However, resilience is a difficult concept to communicate about because terminology is not used consistently across, or even within, domains. This presents a challenge for designers wishing to elicit or understand stakeholders’ requirements for the systems that they are concerned with. To address this, we conducted a workshop with stakeholders working in different areas of academia, industry and policy who are concerned with the resilience of socio-technical systems. The aim of this workshop was to identify what stakeholders might want to convey about resilience and what would help them to communicate effectively. We identified three main characteristics of resilience and three system features that are critical to communication about resilience. These are all illustrated with a diagrammatic framework that was developed from real system examples given by the participants. From the data we propose a set of distinctions that offer a starting point for discussions about resilience with diverse stakeholders.