Price optimization system

• US 7,523,047 B1
• Filed: 12/20/2000
• Issued: 04/21/2009
• Est. Priority Date: 12/20/2000
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: DemandTec Inc. (International Business Machines Corporation)
• Inventors: Venkatraman, Krishna, Lee, Hau, Delurgio, Phil, Neal, Michael, Valentine, Suzanne
• Primary Examiner(s): ALLEN, AKIBA KANELLE
• Application Number:
  US09/741,958
• Time in Patent Office: 3,044 Days
• Field of Search: 705/10, 705/7, 705/1
• US Class Current: 705/7.35
• CPC Class Codes: G06Q 30/0206 Price or cost determination based on market factors
• CPC Class Codes: G06Q 30/0283 Price estimation or determination
• CPC Class Codes: G06Q 30/06 Buying, selling or leasing transactions

Abstract

A method for determining a preferred set of prices for a plurality of products is provided. Generally, a sales model is created. A cost model is also created. A preferred set of prices for the plurality of products based on the sales model and cost model is then generated.

An apparatus for computing a preferred set of prices for a plurality of products is provided. An econometric engine for modeling sales as a function of price to create a sales model is provided. A financial model engine for modeling costs to create a cost model is provided. An optimization engine is coupled to the econometric engine and financial model engine to receive input from the econometric engine and financial model engine, wherein the optimization engine generates the preferred set of prices.

Claims

1. A computer program embedded on a computer readable medium configured to execute steps for computing a preferred set of prices for a plurality of products, comprising:

modeling internal sales, using an econometric engine, as a function of price to create an internal sales model, wherein said econometric engine clusters said plurality of products into discrete sets of related products whereby each said set is made up of highly substitutable related products, further wherein said each said set is defined by a user such that each said set is unique to said user, and wherein said sets are generated by comparing product attribute information, and wherein the internal sales model is for the discrete sets of highly substitutable products;

modeling costs, using an financial model engine, to create a cost model which includes an activity-based costing module configured to receive variable costs and fixed costs, wherein variable costs are related to the volume of sales and wherein fixed costs are independent of volume of sales, wherein said cost model determines a total cost for each product in a given demand group in a given store for a given time period by computing a cost for each selected costing activity including labor, stocking time, transportation, receiving, inventory, bagging, checkout and invoicing, and wherein computing costs utilizes industry data to provide standard cost estimates, and wherein the total cost for each product is dependent upon the attribute information of that product; and

receiving input using an optimization engine coupled to the econometric engine and financial model engine, from the econometric engine and financial model engine, wherein the optimization engine generates the preferred set of prices.

2. The computer program, as recited in claim 1, wherein receiving input using an optimization engine comprises:

storing a plurality of price limiting strategic considerations using a strategy implementation module; and

determining the preferred set of prices using a price calculator connected to the strategy implementation module, the financial model engine, and the econometric engine, based on price limiting strategic considerations, the sales model, and the cost model, further wherein said price limiting strategic considerations iteratively constrain the preferred set of prices to fall within limits conforming to business strategy, wherein said strategy implementation module is a natural language based rules engine which translates said price limiting strategic considerations into rules used by said price calculator.

3. The computer program, as recited in claim 2, further comprising receiving the preferred set of prices using a support tool connected to the optimization engine wherein the support tool from the optimization engine and provides a user interface to a client, wherein the user interface provides the preferred set of prices to the client.

14. The computer program of claim 1 further comprising providing using an imputed variable generator, imputed econometric variables to the optimization engine, the imputed econometric variables used for estimating missing or incomplete product parameters.

4. A computer-implemented method for determining a preferred set of prices for a plurality of products, comprising:

creating an internal sales model, wherein said internal sales model clusters said plurality of products into discrete sets of related products whereby each said set is made up of highly substitutable related products, further wherein said each said set is defined by a user such that each said set is unique to said user, and wherein said sets are generated by comparing product attribute information and wherein the internal sales model is for the discrete sets of highly substitutable products;

creating a cost model which includes activity-based costing, the activity-based costing including fixed costs and variable costs, wherein variable costs are related to the volume of sales and wherein fixed costs are independent of volume of sales, wherein said cost model determines a total cost for each product in a given demand group in a given store for a given time period by computing a cost for each selected costing activity including labor, stocking time, transportation, receiving, inventory, bagging, checkout and invoicing, and wherein computing costs utilizes industry standards data to provide standard cost estimates, and wherein the total cost for each product is dependent upon the attribute information of that product; and

generating the preferred set of prices for the plurality of products based on the internal sales model and cost model.

5. The method, as recited in claim 4, wherein the creating of the internal sales model comprises:

creating a plurality of discrete sets of related products whereby each said set is made up of highly substitutable related products, wherein said sets are generated by comparing product attribute information, and further wherein each discrete sets of related products is a set of at least one product and wherein at least one of the discrete sets of related products is a set of at least two products;

creating an internal sales model for each discrete sets of related products for modeling sales of each discrete sets of related products for a given time period in a given store; and

creating a model for determining the fraction of internal sales of each discrete set of related products made up by each product for said time period.

15. The method of claim 4 further comprising:

receiving raw econometric data;

detecting product parameter inconsistencies in the raw econometric data;

correcting the detected inconsistencies in the raw econometric data to generate a cleansed initial dataset; and

generating imputed econometric variables using the cleansed initial dataset.

6. A computer program embedded on a computer readable medium configured to execute steps for computing a preferred set of prices for a plurality of products comprising:

modeling internal sales, using an econometric engine, as a function of price to create an internal sales model based on Bayesian modeling, wherein data from at least two stores is combined to obtain a Bayesian estimation of the internal sales model, further wherein said econometric engine clusters said plurality of products into discrete sets of related products whereby each said set is made up of highly substitutable related products, wherein said sets are generated by comparing product attribute information, further wherein said each said set is defined by a user such that each said set is unique to said user, and wherein the internal sales model is for the discrete sets of highly substitutable products;

modeling costs, using an financial model engine, to create a cost model which includes an activity-based costing module configured to receive variable costs and fixed costs, wherein variable costs are related to the volume of sales and wherein fixed costs are independent of volume of sales, wherein said cost model determines a total cost for each product in a given demand group in a given store for a given time period by computing a cost for each selected costing activity including labor, stocking time, transportation, receiving, inventory, bagging, checkout and invoicing, and wherein computing costs utilizes industry data to provide standard cost estimates, and wherein the total cost for each product is dependent upon the attribute information of that product; and

receiving input using an optimization engine coupled to the econometric engine and financial model engine, from the econometric engine and financial model engine, wherein the optimization engine generates the preferred set of prices.

7. The computer program of claim 6 wherein the Bayesian model is a Bayesian Shrinkage model.

8. The computer program of claim 7 where the Bayesian Shrinkage model is a multi-stage model.

9. The computer program of claim 6 further comprising providing demand coefficients to the optimization engine, the demand coefficients used for estimating demand given market conditions.

10. The computer program of claim 8 wherein the market conditions include a price point.

11. The computer program of claim 6 further comprising dividing the plurality of products into a plurality of discrete sets of related products whereby each said set is made up of highly substitutable related products, further where at least one of said discrete sets of related products is a set of at least one product and wherein at least one of the discrete sets of related products is a set of at least two products.

12. The computer program of claim 11 further comprising generating a model for determining the fraction of internal sales of each discrete set of related products made up by each product.

13. The computer program of claim 12, further comprising determining an internal sales model for each discrete set of related products so that the optimization engine is able to calculate demand for said products by multiplying the fraction of sales of each discrete set of related products made up by each said product with the internal sales model for the discrete set of related products to which the product belongs.

16. The computer program of claim 6 further comprising providing using an imputed variable generator, imputed econometric variables to the optimization engine, the imputed econometric variables used for estimating missing or incomplete product parameters.