A tsunami hits a densely populated coast. Aid organizations mobilize, and the world watches as several national members of the same global network respond independently in an uncoordinated way. A new treaty is being developed that would allow countries to claim carbon offsets through forest conservation. But national leaders in the same global nonprofit network disagree about its value. A large organization solicits funds from a major U.S. foundation. At the same time, its sister nonprofit, with the same brand name, approaches the foundation. The funder's leaders are confused, and wonder about the seeming conflict. The question of how much to centralize -- or decentralize -- decision-making and operations has dogged global organizations for centuries. Studies of for-profits show that the best answer can be different at different points in an organization's growth. But few such studies exist for nongovernmental organizations (NGOs), and for too many of these nonprofits, the scenarios described above will sound all too familiar because they reflect flashpoints that occur when an operational structure is no longer optimal. What's needed is a way for an NGO's leaders to get out ahead of these flashpoints when possible, by learning to determine in advance when their organization's approach to operations and decision-making need to be revised, and along what lines. In an effort to help with this important task, we synthesized what we've learned through case work with a diverse group of global nonprofits; we also conducted interviews with the leaders and staff at more than 30 global NGOs. Our findings illuminate an emerging approach that blends the best of efficiencies at an organization's center and local innovation in the field.