Smallholder farming systems are critical to agricultural productivity in Ghana, particularly for food security and poverty reduction. However, understanding their cost-effectiveness is complex due to varying resource availability and management practices. A randomized controlled trial (RCT) was conducted among 200 randomly selected smallholder farmers. The trial included three treatment groups: traditional farming practices, improved agricultural technologies, and a combination of both. Data collection encompassed farm inputs, outputs, and costs over a one-year period. The analysis revealed that the combination of traditional farming practices with improved technologies led to an average increase in crop yields by 15% compared to traditional methods alone, while maintaining comparable input costs (standard error ±2%). The randomized field trial provided robust evidence on the cost-effectiveness of different farm management strategies. Policy makers should prioritise interventions that combine both traditional and modern agricultural practices to enhance productivity sustainably. Model estimation used $\hat{\theta}=argmin_{\theta}\sum_i\ell(y_i,f_\theta(x_i))+\lambda\lVert\theta\rVert_2^2$, with performance evaluated using out-of-sample error.