Modeling the Customer Journey

Overview

Once the customer journey has been mapped, it is often a good idea to build a statistical model to interpret the customer journey. That is, there will be points in the journey where the customer must make a choice, a typical example is the customer choosing to proceed with the sales process and to move from one funnel to the next.

The statistical model would then estimate the probabilities that the customer moves from each decision point in the journey to the next. In the case of the pre-sales funnels, the model would give the probability that the customer buys the product after first entering a given funnel.

Capturing the Data

A simple way to capture data about the customer journey is to employ a Customer Relationship Management (CRM) system, which usually allows sales people to set a customer journey and to track a customer's progress. Once the customer data has been captured in a CRM, its data may need to be made available for integration with the davinci platform in order to analyze with davinci's stsatistical tools.

Transition Probabilities

Typical customer journey models are based on Markov Chains and utilize transition matrices.