Summarizing the Mental Customer Journey

Abstract

How do consumers summarize and act on their experiences, as when deciding whether an interaction with a firm was satisfying and whether to buy from it? Previous work on the summary of continuous experiences has tended to focus on a handful of experience patterns and features of those patterns, such as the area under the curve, peak value, and end value. Here, we consider a wider array of possible experience patterns and features of those patterns (27 patterns and 21 features) and quantitatively assess which features are most tied to consumer satisfaction and choice. Contrasting with theories that say fluctuating journeys are most effective, we find that consumers are most satisfied by journeys that stay consistently positive or improve over time, especially toward the experience’s end. Furthermore, we find that several features of the experience predict these outcomes. These features include the end value, slope, integral, peak, and a sentiment score of the words people use to describe the experiences, although the consistently best predictor is the end value. The findings have theoretical implications for summarization, architecting customer journeys, and predicting the success of content, as well as practical implications for return on investment in customer experience optimization.

Keywords: customer experience, customer journey, customer satisfaction, natural language processing, summarization