ABSTRACT: While it is widely recognised that FAIR data means rich metadata, describing data effectively in practice remains a challenge. This applies both to administrative metadata, such as clearly identifying roles like author, producer or depositor, and to descriptive metadata, including titles, descriptions and notes. In this session, we explore what makes a dataset description clear, informative and reusable. What constitutes a meaningful title? Which elements help others to assess the relevance and quality of a dataset? Drawing on ongoing work at DANS, we present guidance on creating richer metadata and invite participants to reflect on their own practices. Presented as a workshop at DANS Open Day 2026 (The Hague)