In detail, the task was to:

  • Identify drivers and triggers within the overall funnel logic for the purchase of an Audi vehicle
  • Take internal and external data points into account to explain the connection between KPIs and each role within the overall framework
  • Show the effect of KPIs within a customer journey from a time perspective


The challenges from a data-sourcing perspective were:

  • Collecting all relevant internal data (based on briefings for the Audi data science team) and providing relevant external data, including online search behavior and pre-processed buzz data
  • Bringing structured and unstructured data into a standardized format for further analysis
  • Identifying a format to investigate online and offline KPIs, with a concentration on time-series analyses
  • Covering spurious correlations based on non-stationary time series, attribution problems due to highly correlated factors (multicollinearity), and time differences between the company’s actions and effects on touchpoints and sales
  • Using statistical methods, machine learning, and probability theory methodologies
  • Enabling an in-depth investigation of customer actions across the entire customer journey


Based on time-series analysis, an exemplary customer journey was created that included:

  • Four major customer touchpoint KPI groups
  • A clustering into the phases: awareness, information/consideration, and purchase, which reflect the different stages of customer involvement
  • Detailed insights into the KPI impact over time
  • A framework which evaluates the impact of performance indicators that clearly affect the company’s business, and mirrors the (marketing) value chain from input to output to outcome