# Node Description

# Differentiation Vertical

The Differentiation Vertical node is designed to take a list of Features, along with an optional list of Variations, and quantify the Vertical Differentiation between each. The quantified Vertical Differentiation between all of the Feature Variations is expressed as a Mean, Standard Deviation (SD), and optional supplier Cost. The Mean and Standard Deviation (SD) are then combined with the Horizontal Differentiation Correlation Matrix in a downstream node to generate a set of part-worth Customer Distributions for each of the Feature Variations.

When Features (or Products) can be rank ordered in an objective way then they are said to exhibit Vertical Differentiation. Features often simultaneously exhibit both Vertical Differentiation and Horizontal Differentiation. But when all Customers universally agree that one Feature is better than another then Vertical Differentiation dominates. In that case, Customers can still disagree as to how much better the one Feature is from the other. This disagreement is reflected in a Customer Distribution comprising a range of part-worth values. When the Customer Distribution is also a Normal Distribution, then the range can be precisely described using a Mean and Standard Deviation (SD).

For example, there is only Vertical Differentiation between a ‘1-year warranty’ and a ‘2-year warranty’ because all Customers universally agree that 2-years is better than 1-year. But that is not to say that all Customers will make the same purchase decision when deciding between the two. Nor is it to say that Price is the only decision factor: Mean and Standard Deviation (SD) of the Customer Distribution, along with Price are all important.

Consider a Market in which Customers generally believe that if a Product fails then it will fail quickly. The Mean of the ‘2-year warranty’ will still be greater than the Mean of the ‘1-year warranty’, but the Standard Deviation (SD) for the ‘2-year warranty’ will shrink. Hence there is a bias to select the ‘1-year warranty’ even among Customers who place a high value on warranties.

Strictly speaking, supply Cost is not part of Vertical Differentiation. Nor, in fact, is Product Price. But the calculation of Feature supply Cost has been included here as a convenience as there is often a relationship between the value of a Feature and the Cost to supply that Feature (the difference between Value and Cost is called ‘Value Created’). But there are many ways to calculate supply Cost, and calculating Feature Cost is but one way. Costs can be calculated at a Store-level, a Product-level, a Feature-level, or a Customer-level. Costs are only important to the Market Simulation when maximizing Profitability – hence Costs need not be calculated at all.

*This Community Node documentation assumes you have already downloaded the open-source KNIME analytics platform and installed the free Market Simulation (Community Edition) plugin. If not, start by returning to Getting Started.*

# Downloads

# #1 Linear Means

## Inputs

#### Related Features

The collection of related Feature names. These may be ordinal Features related by the fact that they can be ranked. For example, the Feature List may be ‘5-star’, ‘4-star’, ‘3-star’, and ‘2-star’. Or these may be categorical Features: for example, ‘Japanese’, ‘Korean’, and ‘German’.

## Node

#### Configuration

Features provide Customers with part-worth value. The Customer Distribution for the Feature collects together all of these part-worth values across all Customers. The Mean of the Input Related Features can each be set according to the Mean Change Method. By default, the Mean will decrease (or increase) in a linear fashion from the Starting Mean to the Ending Mean.

#### SD & Cost

The Standard Deviation (SD) and Cost of each Feature can also be controlled in the Configuration Dialog. These can follow their own path, or be set relative to the Mean.

## Outputs

#### Feature List

The Output Feature List from the node is the output set of objective metrics for each Feature Variation. These Vertical Differentiation objective metrics are combined with the Horizontal Differentiation Correlation Matrix in downstream nodes to generate a set of part-worth Customer Distributions and, ultimately, a Willingness To Pay (WTP) Matrix.

# #2 Individual Means

## Inputs

#### Related Features

The optional ‘Mean’ and ‘SD’ columns in the Input Related Features table can override the default settings in the node Configuration Dialog.

## Node

#### Configuration

The default values from the previous (#1) node Configuration Dialog have not changed.

# #3 Asymptote Means

## Node

#### Configuration

The Means of the list if Related Features are configured to sharply decline from 100.0 then asymptote as the Features get closer to 0.0.

# #4 Feature Variations

## Inputs

#### Features

The list of Input Related Features.

#### Variations

A Variation of a Feature may be associated with a Brand, Product, Channel, Demographic, or Technology. For example, if the Variation is the name of a Brand, then all Products having the same Brand will exhibit the same Variation on the Feature.

## Node

#### Configuration

Change the Feature Means in a linear fashion.

#### Format

When there are Feature Variations added to the second input port, the output format of the Features can be set.

# #5 Customer Distribution

## Inputs

#### Customer ID’s

Upstream nodes generate a list of Customer ID’s.

#### Configuration

An upstream ‘Customer Distributions’ node can do this by generating a list of 10,000 Customers. In fact, the ‘Distribution’ data column will be ignored and the ‘Row ID’ name column will be used.

## Node

#### Configuration

The ‘Mean’ column is configured to contain the Output Customer Distribution.

#### X-Axis

The Standard Deviation (SD) column has been configured to generate the X-Axis data for the downstream ‘Scatter Plot’ node, with Customer ID’s from 1 to 10,000.