Market Simulation

Select Product Line Extension

This Market Simulation is designed to automatically iterate through a number of possible new Products and select the best new Product that is most likely to drive Market Share, Revenue, and Profitability.

Fictitious Scenario: Coca Cola wish to add a Product Line Extension to their existing range of Coke Products. They have already decided to add a 6×330 multi-pack, but haven’t yet selected Bottles or Cans, nor whether to add another Regular flavored cola or a Diet cola. Furthermore, Coke have four new flavors they wish to test: (a) Strawberry, (b) Blackberry, (c) Hazelnut, and (d) Mocha.

The results of the Market Simulation predicts that a new Product comprising of [Regular.Coke + Can + Flavoring.Hazelnut] would increase total profitability of the Coca Cola Product Line the most (after taking into account same Brand Cannibalization).

See also the similar Market Simulation: MS-111 New Product Cannibalization

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.

#1 Feature Selection

Coke are trying to decide which combination of three Features to add to their new Product:

  • Regular vs Diet
  • Can vs Bottle
  • Strawberry, Blackberry, Hazelnut, or Mocha Flavoring

A series of workflow loops allow each Feature combination to be tested. The ‘Table Creator’ nodes define each of the Feature options, while the ‘Table Row To Variable Loop Start’ node iterates through each option.

The rest of this workflow branch adds the set of pre-determined Features and calculates a Product Price and per-unit Cost (the Cost is used to calculate the Profitability of the new Product).

Finally, the ‘Feature Table To List’ node completes the Feature definition so the new Product can be generated in the ‘Product Generator’ node.

Note that the part-worth Customer Distribution values for the new flavoring options are not determined here. It is assumed that the results of outside research are imported into this Market Simulation.

See also: CN-132 Feature Table to List Node

Loop Input #1
Regular vs Diet

Loop Input #2
Can vs Bottle

Input #3
New Flavor

Input #1
Feature Table

Output #1
Feature List

#2 Simulate New Market

The Simulate Market node takes the Input Product Array and the Input Willingness To Pay (WTP) Matrix with the new test Coke Product. The node then predicts how many Customers will buy the new Product and the purchase decision of each Virtual Customer. This future Market prediction (with the new Product) can then be compared to the current Market results (without the new Product).

The results from each combination of Features is collected together and sorted by Profitability. The final results indicate that a new Can of Regular Hazelnut Coke would be the most Profitable addition to Coca Cola’s existing Product Line.

See also: CN-141 Simulate Market Node

Input #1
Product Array

Input #2
WTP Matrix

Output #1
Product Array

Results #1
Final