Passive metering provides a direct access to consumers’ online behavior. This wealth of unfiltered data is rife with opportunities but it also challenges traditional data analysis, because it is novel in kind and big in volume. In this workshop, I will focus on the analysis of a specific sort of such data, namely online customer journeys. What do customers do (online) before they buy (online)? What are the typical paths which lead to the online purchase of a given category of products? Can these paths be made sense of purely on the basis of online behavior? This is not completely new: weblogs of commercial websites, online touchpoints and referral data have been used to provide valuable insights into purchase paths. But access to the entirety of navigation data offers new perspectives and new challenges, because the nature of events is open-ended – all urls might be on the way, all paths are open. My aim will be to present and discuss the various strategies that are available to tackle this wilderness, both in terms of which data is used and which analyzing techniques are appealed to. Regarding data, salient issues concern the definition of the journey (e.g. how far back before the purchase?) and the amount of pre-treatment (website categorization). Regarding analysis, the main options concern the reckoning of time (flat analysis versus sequences or timed series) and navigation (history dependent or history free paths), and model free or model based analysis (e.g. in terms of how influence, information and action get combined). We will consider in particular the application of traditional clustering techniques, sequence based clustering and process mining.