This paper aims at refining and suggesting new leads in the research on international expansion patterns. The model developed by Uppsala scholars (Johanson & Vahlne, 1977; Johanson & Wiedersheim-Paul, 1972) fails to account for patterns observed empirically. However their reasoning in terms of increasing psychic distance is appealing and makes sense logically: managers being risk averse, they will favor locations they feel familiar with in order to reduce the uncertainty associated with doing business abroad. Therefore we suggest new approaches to be explored: the first one is a new operationalization of the concept of distance, since the currently dominant one, the Kogut & Singh (1988) index is flawed; a second one discusses the potential of adding the notion of ambidexterity, from the exploration/exploitation paradigm, into the picture; a last one confronts neoinstitutionalism and game theory to predict competitors moves and their influence on the location choices of the firm. We hope to contribute to the international business literature in two ways: first, validating and completing the learning perspective of internationalization; second, bringing evidence that the operationalization of distance needs to be modified to be useful.