Article dans une revue: The relevance of school interventions in ecological contexts is hard to anticipate, because of the complexity of school and social environments. It is thus important to test rigorously interventions at scale using randomized field experiments. I will illustrate how simple parental interventions designed by local French education authorities were tested at a large scale. These low cost interventions consist in collective meetings between parents and school staff and they have significant effects on behavior and track choices. Results from these experiments have been influential at encouraging action towards parents in school. I will discuss how the current literature is attentive to external validity issues that condition how much a given experiment is informative outside its place of intervention.