We analyze the effects of four randomized social experiments, involving early and intensive active labour market policy, conducted in Denmark in 2008. The experiments entailed different combinations of early and intensive treatment in terms of meetings and active labour market programmes. The effects are remarkable; frequent meetings between newly unemployed workers and case workers can increase employment rates over the next two years by up to 5 weeks, corresponding to 10%. For men, we find evidence of a threat effect of having to participate in early active labour market programmes, while no such effect is found for women. In general, we find large differences between men and women, especially in the exact timing of the effects. We conduct a cost-benefit analysis of each of the four experiments and find that meetings yield the largest net benefits.