User engagement plays a central role in a project such as NFDI4Earth, which seeks, among other objectives, to foster FAIR and open data practices in the Earth System Sciences (ESS). To support this goal, NFDI4Earth organizes a range of community-oriented activities that facilitate dialogue, collect feedback, and address the needs of researchers at different career stages and from diverse subcommunities. This report highlights three such events, spanning local, national, and international contexts. The NFDI4Earth Academy regularly hosts free and open online Coffee Lectures addressing key topics in Research Data Management (RDM) within the ESS. While the lectures are primarily aimed at Academy Fellows, particularly early-career PhDs and postdocs, they are also accessible to the broader community. The Coffee Lecture held in February 2025 focused on the NFDI4Earth OneStop4All portal and included a presentation and an (open) exchange. A further community engagement activity was the Faculty Day Workshop held in May 2025, which brought together members of the environmental faculty at TU Dresden across different career stages. The workshop provided an opportunity to present the work of the local NFDI team and to discuss a broader range of topics relevant to research data management. The yearly European Geosciences Union (EGU) General Assembly (GA) attracts more than 20,000 international scientists. In 2026, the first NFDI4Earth short course on data discovery, repositories, and learning materials was successfully held. Through a mix of interactive discussions, live demonstrations of OneStop4All, and open discussions, we gained insights into current RDM challenges and needs, relevant data sources, and other valuable input for the further developments. Together, these activities provided insights into researchers’ data practices, RDM priorities, and practical challenges across different career stages and disciplinary contexts. The results reveal a fragmented data landscape. Participants referred to a wide range of disciplinary and institutional repositories, governmental data portals, research infrastructures, and general-purpose platforms. Across the three events, Copernicus/ESA Sentinel Hub, Zenodo, national services and repositories, and PANGAEA were among the most frequently mentioned sources. This diversity underlines the continuing importance of effective data discovery and guidance in selecting suitable repositories. The most prominent RDM needs relate to metadata, FAIR principles, data publishing and repositories, data discovery and access, interoperability, and practical support across the research data lifecycle. Participants highlighted difficulties with incomplete or non-standardised metadata, integrating heterogeneous datasets, and accessing and working with large datasets. The reported challenges span the entire data lifecycle, with large datasets, metadata quality and standardisation, data integration and interoperability, discoverability, and limited time, skills, and practical support emerging as recurring concerns. Emerging requirements, such as making datasets AI/ML-ready, further increase the need for practical and scalable solutions. The findings demonstrate that researchers need not only information and training, but also clear guidance, appropriate infrastructure, and accessible tools that fit into everyday workflows.