Can an implementation-driven analysis of project success be used as a more granular instrument for assessing the effectiveness of World Bank project By focusing on how projects perform, this paper attempts to capture variation hitherto unexplored in the aid effectiveness literature. This offers greater precision for diagnosing implementation challenges throughout the project cycle, producing a cross-cutting instrument that reaches across country-, time-, and sector- based approaches. Using data from the Global Delivery Initiative's "Delivery Challenges in Operations for Development Effectiveness" database and indicators from more than 5,000 lending projects (1995 - 2015), the paper examines project performance and the achievement of development objectives across 42 specific delivery challenges. Bayesian model averaging is used for a holistic assessment of the relative impacts of each challenge alongside a battery of structural and contextual covariates. The findings show that issues of project design, ineffective monitoring, and weak organizational capacity have systematically hindered the World Bank's performance and achievement of the indicators. Conversely, while financial instability and weaknesses in stakeholder engagement can hinder success, their identification and treatment ultimately improves project performance.