A strong case has been made for freely available, high quality data on species occurrence, in order to track changes in biodiversity. However, one of the main issues surrounding the provision of such data is that sources vary in quality, scope, and accuracy. Therefore publishers of such data must face the challenge of maximizing quality, utility and breadth of data coverage, in order to make such data useful to users. Here, we report a number of recommendations that stem from a content need assessment survey conducted by the Global Biodiversity Information Facility (GBIF). Through this survey, we aimed to distil the main user needs regarding biodiversity data. We find a broad range of recommendations from the survey respondents, principally concerning issues such as data quality, bias, and coverage, and extending ease of access. We recommend a candidate set of actions for the GBIF that fall into three classes: 1) addressing data gaps, data volume, and data quality, 2) aggregating new kinds of data for new applications, and 3) promoting ease-of-use and providing incentives for wider use. Addressing the challenge of providing high quality primary biodiversity data can potentially serve the needs of many international biodiversity initiatives, including the new 2020 biodiversity targets of the Convention on Biological Diversity, the emerging global biodiversity observation network (GEO BON), and the new Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (IPBES).