Ecosystems degradation, and consequently biodiversity loss, has severe impacts on people around the world. The Intergovernmental Platform on Biodiversity and Ecosystem Services (IPBES) is one of the international initiatives that have emerged to inform policy makers and aid decisions to prevent further global biodiversity loss, focusing on the interdependence between natural systems and human culture. IPBES promotes the use of scenarios and modelling approaches as a fundamental tool to advance the understanding of the relationships between drivers of change, Nature’s Contributions to People (NCP), and social systems. Local-scale case studies with a system approach demonstrating how current knowledge can be used to inform decision-making are still scarce. Here, we present a comprehensive conceptual model and a series of four scenarios under different policies for shea tree species management, as a case-study of applying systems thinking and the NCP concept to a local-scale socio-ecological system. We first characterized the central processes, NCP, drivers and pressures affecting the shea tree system, to investigate the impacts of the multiple uses of the shea tree species on the system as a whole. We then described potential policy options, developed four scenarios, and evaluated them by a Bayesian Belief Network (BBN). We predicted qualitative outcomes of the proposed scenarios: Business-as-usual (BAU), “Conservation and fair trade”, “Agroforestry and fair trade” and “Industrial development”. We found that the scenarios focussing on conservation, fair trade and agroforestry, can improve the conservation status of shea trees, and enhance wellbeing in the local communities. In this case study, we demonstrate that the development of a comprehensive conceptual model at a local scale can be a useful exercise to identify opportunities for effective policy strategies and social innovation. The shea tree case study can provide an example for modelling non-timber forest products in other regions around the world that face similar drivers and pressures. Species for which this model could be adapted include Central and South American species such as the Brazilian nut (Bertholletia excelsa), cocoa (Theobroma cacao), andiroba (Carapa guianensis), açai (Euterpe oleracea) and the wax palm (Ceroxylon quindiuense). The model and workflow applied here may thus be used to understand similar socio-ecological systems with local and international economic value across the Neotropical region.