Although tags play a critical role in many social media, their link to the geographic distribution of user generated videos has been little investigated. In this paper, we ana- lyze the correlation between the geographic distribution of a video’s views and the tags attached to this video in a Youtube dataset. We show that tags can be interpreted as markers of a video’s geographic diffusion, with some tags strongly linked to well identified geographic areas. Based on our findings, we explore whether the distribution of a video’s views can be predicted from its tags. We demon- strate how this predictive power could help improve on-line video services by preferentially storing videos close to where they are likely to be viewed. Our results show that even with a simplistic approach we are able to predict a minimum of 65.9% of a video’s views for a majority of videos, and that a tag-based placement strategy can improve the hit rate of a distributed on-line video service by up to 6.8% globally, with an improvement of up to 34% in the USA.