Calls for more ‘adaptive programming’ have been prominent in international development practice for over a decade. Learning-by-doing is a crucial element of this, but programmes have often found it challenging to become more learning oriented. Establishing some form of reflective practice, against countervailing incentives, is difficult. Incorporating data collection processes that generate useful, timely and practical information to inform these reflections is even more so.

This paper explores MUVA - an adaptive female economic empowerment programme in Mozambique. MUVA, we suggest, is atypically evidence-led. It combines systematic, inclusive reflective practice with extensive real-time data collection. We describe the fundamental features of MUVA’s monitoring, evaluation and learning (MEL) approach that supported this. One, how data collection and analysis are synchronised with set cycles for learning and adapting projects. Two, how MEL systems are designed to prioritise actionable learning, with data collection oriented more to the needs of implementing staff than to the reporting requirements of funders.

This approach was enabled by building collective ownership over the programme’s objectives and the purpose of MEL from the outset. Implementers are asked about their motivations, and these are related to the programme’s Theory of Change. The evidence culture is supported by the proximity of MEL staff to implementing staff; and through structuring upwards accountability to funders around justifying evidence-based adaptations instead of reporting on more narrow indicators. We conclude by considering the relevance, or not, of MUVA’s approach to programmes in other contexts or issue areas trying to replicate a similarly evidence-informed approach to adaptive management.