I love the work of GiveWell in attempting to rigorously evaluate charities to help donors pick the most evidence-supported charitable organizations. However, this assumes that donors will give a set amount of money and are just looking for a recipient. In practice, I think donors are more likely to give to some groups that tug at the heart strings, that are related to current news, etc. Some charities won't be supported by rich governments for ideological reasons. So besides using metrics like lives saved per dollar, how can we maximize donations from rich countries and ensure that they do the most good?