Statistics are often used to reveal significant differences between online and campus-based education. The existence of online courses with low completion rates is often used to justify the inherent inferiority of online education compared to traditional classroom teaching. Our study revealed that this type of conclusion has little substance. We have performed three closely linked analyses of empirical data from Linnaeus University aimed at reaching a better understanding of completion rates. Differences in completion rates revealed themselves to be more substantial between faculties than between distribution forms. The key-factor lies in design. Courses with the highest completion rates had three things in common; active discussion forums, complementing media and collaborative activities. We believe that the time has come to move away from theoretical models of learning where web-based learning/distance learning/e-learning are seen as simply emphasizing the separation of teacher and students. Low completion rates should instead be addressed as a lack of insight and respect for the consequences of online pedagogical practice and its prerequisites.