Abstract. In this paper, we present a system that facilitates the analysis of user activity within a web video. Previous research in user-based techniques has assumed an extra effort from the users, such as video replies, comments, tags, and annotations. We have developed and evaluated the SocialSkip system, which improves sense making of web videos by visualizing the simplest form of user interactions with video, such as pause, and seek. In contrast to previous stand-alone implementations, the SocialSkip system employs a web-video player and cloud-based resources (application logic, database, content). The system was validated with two user studies, which provided several hundreds of user interactions with five types of web video (sports, comedy, lecture, documentary, how-to). We found that seeking activity within web video is reversely proportional to how interesting the video is. Moreover, we suggest that if the users are actively seeking for information within a video (e.g., lecture, how-to), then the user activity graph could be associated with the semantics of the video. Finally, SocialSkip provides an open architecture for collaborative contributions to the analysis of the user activity data, in a fashion similar to the TRECVID workshop series.