Case Study

SUV Features and Products

This Case Study provides a Market Simulation of the SUV Market. Feature research data comes from J.D. Power. Sales data comes from the Chinese SUV market in 2013.

This Market Simulation tunes 12 differentiating Features across 15 SUV vehicles:

Midsize SUV Vehicles Differentiating Features
Audi Q5 Engine Power
BYD S6 Utility
Chery Tiggo Brand
DFM Jingyi Safety
DFM Lingzhi Miles per Gallon
Great Wall Hover Instrumentation
Honda CR-V Style
Hyundai ix35 Performance
Kia Zhipao Comfort
Mercedes MB GLK Body and Interior
Nissan Qashqai Powertrain
Roewe W5 Accessories
Toyota Highlander
Toyota RAV4
VW Tiguan


This Case Study provides a high-level overview of the workflow without detailed explanation. It assumes you are already somewhat familiar with KNIME and Market Simulation. If not, start by reviewing the Building Blocks and Community Nodes.

Differentiation Pattern

This Case Study again uses a very common Differentiation Workflow Pattern to build a set of part-worth Customer Distributions for each of the 12 Features that define an SUV. Click here for a detailed review of the Differentiation Workflow Pattern.

#1 Brand

Brand Differentiation

Five Brand levels have been defined in approximate rank order (from most valuable to least valuable):

  • Euro Luxury
  • Euro Standard
  • Japanese
  • Korean
  • Chinese

Brand Variations have then been defined for every SUV in the Market.

The Feature Tuning Loop adjusts the Vertical Differentiation (Starting Mean and Ending Mean), and the Strange Differentiation (SD Change Factor). Tuning Parameters are passed into the workflow as Flow Variables. The Horizontal Differentiation between Brands is not tuned.

#2 Feature Tuning

Tuning Loop Start

Each of the Features are tuned separately, with an outer loop controlling which Feature is being tuned. Brand is tuned first, followed by Horse Power (HP), Miles per Gallon (MPG), etc.

The Tuning Loop Start node gathers all of the Input Parameters related to each Feature and starts adjusting them. For example, for the Accessories Feature (AUX) the Tuning Loop Start node will adjust:

  • AUX_Starting_Mean, and
  • AUX_Ending_Mean_Ratio

The other Input Parameters are then collected together and passed into the workflow as fixed Flow Variables.

#3 End Tuning

Tuning Loop End

After all of the Differentiating Features have been defined, they are aggregated together into Products along with a Customer Willingness To Pay (WTP) Matrix.

The ‘Simulate Market’ node calculates the ‘Quantity Error’ and this is passed to the Tuning Loop End. If the total ‘Quantity Error’ is decreasing then the adjusted Tuning Parameters are retained.