One of the main challenges in distributed development is ensuring effective communication and coordination among the distributed teams. In this context, little attention has been paid so far to coordination in software testing. In distributed development environments, testing is often performed by specialized teams that operate as independent quality assurance centers. The use of these centers can be advantageous from both an economic and a software quality perspective. These benefits, however, are offset by severe difficulties in coordination between testing and software development centers. Test centers operate as isolated silos and have little to no interactions with developers, which can result in multiple problems that lead to poor quality of software. Based on our preliminary investigation, we claim that we need to rethink the way testing is performed in distributed development environments. We then present a possible research agenda that would help address the identified issues and discuss the main challenges involved.