We used Qualitative Comparative Analysis (QCA) to study the combinations of factors that are consistently related to success or failure of grants given to farmer groups. Using data from a sample of 26 grant beneficiaries, we explored whether baseline characteristics of the organisations related to group sales, organisational scale and organisational strength could predict the intended outcomes of the grant system: improved access to markets for member products, increased organisational capacity, and more income to pay organisational expenses. We explain the calibration process used to assign each
organisation to (fuzzy-set) conditions, and the iterative process of QCA to explore the resulting truthtable for plausible causal configurations that may help to target grant funds. We use the ambiguities in the evaluation of success or failure of certain organisations to verify the robustness of the analysis
under real-world conditions of measurement error. We detected some single conditions consistently related with success, especially if they were sourcing raw material from members or the spot market, and could triangulate these patterns with logistic regression. The grants to the older, larger and stronger organisations were consistently unsuccessful, because the grant resulted in under-scaled investments in secondary activities that were discontinued after pilot experiences. Finally, we discuss the strengths and weaknesses of QCA as a method for explorative research and causal inference.