Training can better meet user needs by involving the potential users early in the process. Here, development of open science training for food science students and professionals began with 11 guided interviews of 15 beneficiaries in a European project aiming to build a ‘food cloud’ of datasets and services (FNS-Cloud, H2020 No. 863059). Discussions covered what partners want to learn, how they prefer to learn, and who are their ideal trainers. Inductive coding of interview transcripts with NVivo 12 qualitative analysis software revealed an inclination for technical training with a focus on data. Face-to-face learning and on-demand elearning offered by younger scientists were preferred methods and trainers. Most interviewees also talked about ‘food cloud’-specific fears and desires. These interviews are now the foundation of three well received elearning courses and two workshop series supporting the value of user input in early course development decisions.