When 95% of sales happen in a brick & mortar store but 60% of shoppers complete online research, how do you optimize your online presence? Determining the value of a touchpoint is a common marketing problem with a variety of methodologies used to find a solution. This presentation will explore an application of Markov models to determine the value of a touchpoint throughout the onsite purchase funnel. More sophisticated than traditional first-touch/last-touch rules-based attribution, Markov models provide a way to represent the dependencies of different touches and their values through machine learning. These values can then be used to optimize websites and marketing spend.