A lot of money is wasted in international development, prompting research that identified various systemic and procedural features of the observed development inefficiency. Within procedural inefficiency, problems in project implementation have led to unmet objectives within proposed timeframes and budgets, and consequently, projects ‘failing'. Subsequently, numerous project designs have been dismissed for not producing the expected results and not ‘working'. Yet, the observed failure to achieve project objectives is often a result of inefficiency rather than actual unsuitability of the project approach and design. In this paper, we address the core issue of random sampling and recruitment of project ‘beneficiaries' without identifying whether these partners are suitable during project implementation. Yet, it is these very ‘beneficiaries' whose engagement, practices and performance will be assessed to qualify the project's success. Conventional targeting methods are typically needs-based and fail to address concerns of responsiveness to project objectives, efficiency, and legitimacy. The aim of this paper is to develop, use, and evaluate a group selection tool for purposive project group sampling based on an assumption that their performance could be predicted ex ante. The proposed tool is composed of two multi-indicator categories: group material well-being, and group capacity and agency. To determine how the tool is associated with positive outcomes, we triangulated various project data. Some relevant results from comparing performance data with the adapted group classification include that groups with higher well-being often formed strategic partnerships with external actors, while those with higher agency and lower well-being were more likely to collaborate with each other. Overall, higher agency was correlated with higher overall performance. While the tool needs further refinement, we expect it to contribute positively to procedural inefficiency concerns in development projects and programs, and to lead to better, faster, more sustainable development outcomes.