The Demand Shredder node removes values from the Willingness To Pay (WTP) Matrix to reduce the possibility of Customers switching to unlikely Products outside of their normal Consideration Set.

Many Competitor Stores sell identical Products, and many Brands offer Products that are very similar. However Customers often do not consider these identical or highly similar Products, either because they are unaware of a Competitive Store’s offering, or because they prefer not to investigate Brands they are unfamiliar with.

The Market Simulation Profit Engine assumes, by default, that all Customers are aware of all Product offerings. Therefore it allows a Virtual Customer to select perhaps their least favorite Product if only the Price of that Product were sufficiently discounted. However, this usually does not reflect the reality of the Market. A real-world Customer may never select a least favorite Product simply because they were not even aware of that Product’s existence. A real-world Customer may only ever investigate the Product offerings from a single Store or a single Brand.

The Demand Shredder node can be configured, for example, so that Virtual Customers consider only their Top 3 to 4 Stores, their Top 2 to 5 Brands, and their Top 1 to 3 Locations. Demand Shredder Nodes can be chained together so that, in this example, the first node would be configured to shred less desirable Brands, the second node would be configured to shred less desirable Stores, and the third node would be configured to shred less desirable Locations.

In this way, the Demand Shredder node is able to model the ‘Search Cost’ a Customer would suffer if they were to do extensive research. Most Customers are prepared to forego an excellent deal if it means they can reduce the Search Cost required to identify such a deal. In other words, these Customers prefer a good deal with low Search Costs than the very best deal with high Search Costs.

The Demand Shredder node will always retain the Customer’s first Top Choice. The Top Choice is determined by maximizing the Consumer Surplus the Virtual Customer would receive across all Products in the Market. If the Demand Shredder is shredding Brands, then the Brand of the Customer’s top Product choice is always retained regardless of the Store, Location, or other Product Attribute. The advantage of retaining the Customer’s Top Choice is that the results of an upstream Tuning Algorithm should not be impacted by a Demand Shredder node. If, on the other hand, values were randomly shredded from the WTP Matrix, then a Profit Engine would yield different results between the shredded WTP Matrix and the original unshredded WTP Matrix.

Demand Shredder nodes are best located downstream a Market Simulation Tuning node after the final WTP Matrix values have already been calculated.

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.

Shredded Demand

After the last ‘Shred by Location’ node, there are many zero-value gaps in the WTP Matrix. The remaining Products with non-zero values make up each Customer’s Consideration Set.

Before

The Demand Curve before Demand Shredding is relatively steep. Any Price change will be seen by all Customers, and hence Price Sensitivity will be high.

To maximize Profit, Price needs to be decreased by 17% to $82.49.

After

After Demand Shredding, the Demand Curve is much flatter. Now most Customers will not see the Price change by any given Product. Competition has been reduced and Price Sensitivity is low.

Now to maximize Profit, Price needs to be increased by 5% to $104.99.