Small vs big user studies (and making lots of changes)

Jakob Nielsen points out that more testers result in diminishing returns. The key is acting on feedback from a small number of users regularly.