Demand modeling and forecasting for configurable products or services

• US 8,301,486 B1
• Filed: 09/12/2008
• Issued: 10/30/2012
• Est. Priority Date: 09/14/2007
• Status: Expired due to Fees

• Current Assignee: Emcien, Inc.
• Original Assignee: Emcien, Inc.
• Inventors: Marsten, Roy Earl
• Primary Examiner(s): Dixon, Thomas
• Assistant Examiner(s): Vizvary, Gerald
• Application Number:
  US12/209,429
• Time in Patent Office: 1,509 Days
• Field of Search: 705/7.25, 705/7.28, 705/7.29, 705/7.31, 705/36.R, 1/1, 707/5
• US Class Current: 705/7.31
• CPC Class Codes: G06Q 10/00 Administration; Management
• CPC Class Codes: G06Q 30/00 Commerce

Abstract

A system and method for modeling demand for a configurable product or service and then forecasting demand for said product or service based on the statistical associations of options across features within said product or service. By use of this method a manufacturer, for example, would be able to produce product combinations in advance with an increased probability that they would meet customer demand or project parts inventory needs and build fewer unwanted combinations.

Claims

1. A process for modeling demand for a configurable product or service, comprising the steps of:

(a) providing an input data set to a processor containing a plurality of features, each feature comprising one or more options for the configurable product or service;

(b) defining a plurality of first option-sets with a processor, wherein each first option-set comprises one or more selectable options corresponding to a feature;

(c) defining a first order joint-take-rate with a processor as comprising a combined probability of choosing each available option within a first option-set, defining a second order joint-take-rate with a processor as comprising a combined probability of choosing pairs of options available within two first option-sets;

(d) obtaining a second-order joint-take-rate for pairs of first option-sets with a processor;

(e) determining if correlations exist among the second order joint-take-rates of the pairs of first option-sets with a processor;

(f) if correlations exist among the joint-take-rates for pairs of the first option-sets, then grouping features with a processor corresponding to the related first option sets together to form feature-sets;

(g) calculating an n-order joint-take-rate for each feature-set formed from the grouping with a processor, wherein n corresponds to a number of first option-sets forming a respective feature-set;

(h) creating second option-sets corresponding to each of the plurality of feature-sets with a processor using a weighted random selection that relies on the n-order joint-take rate of a respective feature-set;

(i) combining the second option-sets with a processor to define at least one configurable product or service; and

(j) displaying the second option-sets on an output device.

2. The process of claim 1, further comprising a step of:

creating a rules validator to validate said at least one configurable product or service, wherein steps (h)-(j) are repeated if said configurable product or service is found to be invalid.

3. The process of claim 1, further comprising a step of:

parsing said input data to produce a formatted data input set that allows a calculation of a first order joint-take-rate for each first option-set.

4. The process of claim 1, wherein determining if correlations exist among the second order joint-take-rates of the pairs of first option-sets comprises determining if a chi-square value is greater than a predetermined threshold.

5. The process of claim 1, wherein said first option-sets contain at least two options.

6. The process of claim 1, wherein steps (h)-(j) are executed once for each unit of a product intended to be produced for an inventory.

7. The process of claim 1, further comprises creating a rules validator that bars calculation of invalid second option-sets.

8. The process of claim 6, wherein step (h) is further executed for a multiple of the number of products in said inventory for the purpose of determining outliers in said inventory.

9. The process of claim 8, further including the step of:

abandoning an inventory and substituting alternative into its place based on a determination that said outliers fall outside a predetermined tolerance factor.

10. The process of claim 8, wherein said multiple is at least 3.

11. The process of claim 4, wherein said threshold is automatically determined.

12. The process of claim 4, wherein said threshold is determined through the steps of:

choosing a threshold;

iterating through steps (h) through (j) until a set of configurable products or services is created; and

accepting said set if its number of unique configurations is comparable to the number of unique configurations in said input data set.

13. The process of claim 1, wherein steps (h)-(j) are executed a predetermined number of times.

14. The process of claim 1, wherein using a weighted random selection further comprises using a Monte Carlo methodology.

15. The process of claim 14, wherein the Monte Carlo methodology comprises generating a uniform random number in the range from zero to one, and then using an inverse of a cumulative distribution function.

16. A computer program product comprising a computer usable medium having a computer readable program code embodied therein, said computer readable program code adapted to be executed to implement a method for modeling demand for a configurable product or service, said method comprising:

(a) providing an input data set containing a plurality of features, each feature comprising one or more options for the configurable product or service;

(b) defining a plurality of first option-sets, wherein each first option-set comprises one or more selectable options corresponding to a feature;

(c) defining a first order joint-take-rate as comprising a combined probability of choosing each available option within a first option-set, defining a second order joint-take-rate as comprising a combined probability of choosing pairs of options available within two first option-sets;

(d) obtaining a second-order joint-take-rate for pairs of first option-sets;

(e) determining if correlations exist among the second order joint-take-rates of the pairs of first option-sets;

(f) if correlations exist among the joint-take-rates for pairs of the first option-sets, then grouping features corresponding to the related first option sets together to form feature-sets;

(g) calculating an n-order joint-take-rate for each feature-set formed from the grouping, wherein n corresponds to a number of first option-sets forming a respective feature-set;

(h) creating second option-sets corresponding to each of the plurality of feature-sets using a weighted random selection that relies on the n-order joint-take rate of a respective feature-set;

(i) combining the second option-sets to define at least one configurable product or service; and

(j) displaying the second option-sets on an output device.

17. The computer program product of claim 16, wherein determining if correlations exist among the second order joint-take-rates of the pairs of first option-sets comprises determining if a chi-square value is greater than a predetermined threshold.

18. The computer program product of claim 17, wherein said threshold is automatically determined.