The article analyzes the transition from field experiments to large-scale policy interventions, where many successful programs do not achieve the expected results when scaling up[1][2]. A phenomenon called the "scale-up effect" causes a change in the effect size of an intervention when moving from a small research setting to a population-wide implementation, which can be negative or positive[1]. This effect is influenced by various factors, including the incentives of actors in the knowledge creation market[1][2]. The authors identify three key ingredients for understanding the scale-up effect through an economic lens[1]. They suggest twelve measures for researchers, policy makers, funders and stakeholders to avoid threats to scalability[2]. Suggestions include evaluating experimental results cautiously, avoiding scaling up without sufficient evidence of efficacy, encouraging replication studies, and using multi-site trials to detect variability in effects[1][2]. The aim is to strengthen confidence in experimental findings for their better use in major policies[2].