Sequential data are found in many settings, e.g., as sequences of websites that users visit, as sequences of travel locations, or as sequences of songs a user listened to. To improve the understanding of the mechanisms that underlie the generation of such sequences, we developed a novel approach called HypTrails. It utilizes Bayesian hypothesis testing and rst-order Markov chain models in order to enable the comparison of a set of hypotheses with respect to their plausibility considering some observed data. Each of the hypotheses captures a belief in transitions between the states. It can be derived from theory in the application domain, from other related datasets, or from human intuition. We applied this approach to study several phenomena in the online world, e.g., navigation behavior in Wikipedia or urban mobility data. In this talk, we want to give an introduction to HypTrails [4] and showcase selected real-world applications [3, 1, 2] that utilize it.