Huge sums in billions of dollars have been poured into research capacity strengthening (RCS) over the last two decades, yet the return on this investment in low- and middle-income countries remains frustratingly uneven. Despite more training and better funding, systemic issues, such as weak data practices, poor reproducibility, and the failure to translate findings into policy, persist. We argue that the problem is not a lack of resources, but a flaw in design. Most capacity-building efforts still rely on linear, “pipeline” models that treat research as a series of isolated skills to be learned, rather than an interconnected system to be managed. In this perspective, we introduce the Global Research Lifecycle Ecosystem (GRLE) as a diagnostic framework to identify and mitigate these systemic failures. This framework shifts the focus from individual training to ecosystem performance. Unlike traditional models, the GRLE maps the messy, real-world interactions between people, digital tools, and institutional incentives. We define “Lifecycle Coherence” as a core design principle for Digital Research Infrastructure (DRI), utilizing unified metadata schemas and automated audit trails to ensure data integrity. We show how “ecosystem failures,” like disconnected data governance or perverse promotion criteria, create a cascade of inefficiency that undermines development goals. We also examine the double-edged sword of digital infrastructure and AI: technologies that can either bridge these gaps or widen them, depending on how they are deployed. Through illustrative use cases involving national and global funding agencies (e.g., EDCTP, Global Fund, and the Bill & Melinda Gates Foundation), we conclude with a call to funders and institutions to move beyond fragmented projects and embrace ecosystem stewardship, ensuring research investments actually serve the Sustainable Development Goals and AU Agenda 2063.
Adedeji et al. (Wed,) studied this question.