Inventory and price decision support

• US 6,910,017 B1
• Filed: 03/05/1999
• Issued: 06/21/2005
• Est. Priority Date: 03/05/1999
• Status: Expired due to Term

• Current Assignee: Oracle International Corporation (Oracle Corporation)
• Original Assignee: ProfitLogic, Inc. (Oracle Corporation)
• Inventors: Levy, Michael, Bible, John, Woo, Jonathan
• Primary Examiner(s): Thomas, Joseph
• Assistant Examiner(s): Morgan, Robert W.
• Application Number:
  US09/263,979
• Time in Patent Office: 2,300 Days
• Field of Search: 705/400, 705/10, 705 28- 29
• US Class Current: 705/7.29
• CPC Class Codes: G06Q 10/06 Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
• CPC Class Codes: G06Q 10/0637 Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals
• CPC Class Codes: G06Q 10/06375 Prediction of business process outcome or impact based on a proposed change
• CPC Class Codes: G06Q 10/0875 Itemisation or classification of parts, supplies or services, e.g. bill of materials
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data

Abstract

From data that includes prices and unit sales of an item of commerce for a succession of time periods during which the item was sold, a relationship is derived that represents variation of unit sales of the item as a function of price and time of sale. A maximum profit is predicted by optimizing the price or inventory or both over time in accordance with the relationship. The optimized price or inventory or both over time are used in connection with making price decisions or inventory decisions or both with respect to marketing the item of commerce. A group of the data is aggregated, the group being chosen so that the aggregated data for the group will have an enhanced signal-to-noise ratio with respect to information about sales trends. The aggregated data for the group is analyzed for use in making price decisions with respect to marketing the items of commerce. A series of charts illustrating variations in sales over time are generated automatically, different charts being for different selected sets of the data.

Claims

1. Software stored on a machine-readable medium that configures a machine to do the following:

from data that includes prices and unit sales of an item of commerce for a succession of time periods during which the item was sold, derive a relationship of the form

S(t)=N_o*N_c*t^γ

*[e<sup>−

(t−

tpk)α</sup>

+C]

 

where S(t) represents gross sales revenues of the item;

t is time;

N_o is a curve fitting parameter;

N_c is a demand multiplier representing an increase in sales that occurs when price is reduced;

t_pk is a time of peak sales;

α and

γ

are curve fitting constants;

C is a residual sales rate at large values of t; and

N_c is a function representing a demand multiplier that is a function of price,

predict a maximum profit by optimizing the price or inventory or both over time in accordance with the aforesaid relationship and in accord with a relationship wherein profit is equal to sales revenues minus costs, and

use the optimized price or inventory or both over time in connection with making price decisions or inventory decisions or both with respect to marketing the item of commerce.

2. A system comprising

means for deriving, from data that includes prices and unit sales of an item of commerce for a succession of time periods during which the item was sold, a relationship of the form:

S(t)=N_o*N_c*t^γ

*[e<sup>−

(t−

tpk)α</sup>

+C]

 

where S(t) represents gross sales revenues of the item;

t is time;

N_o is a curve fitting parameter;

N_c is a demand multiplier representing an increase in sales that occurs when price is reduced;

t_pk is a time of peak sales;

α and

γ

are curve fitting constants;

C is a residual sales rate at large values of t; and

N_c is a function representing a demand multiplier that is a function of price,

means for predicting a maximum profit by optimizing the price or inventory or both over time in accordance with the aforesaid relationship and in accord with a relationship wherein profit is equal to sales revenues minus costs, and

means for using the optimized price and sales data over time in connection with making price decisions or inventory decisions or both with respect to marketing the time of commerce.

3. A machine-based method for making price decisions or inventory decisions or both with respect to the marketing of an item of commerce, the method comprising

A. executing, on a machine, software that configures the machine to perform the following steps

i. from data that includes prices and unit sales of the item of commerce for a succession of time periods during which the item was sold, deriving a relationship of the form

S(t)=N_o*N_c*t^γ

*[e<sup>−

(t−

tpk)α</sup>

+C]

 

where S(t) represents gross sales revenues of the item;

t is time;

N_o is a curve fitting parameter;

N_c is a demand multiplier representing an increase in sales that occurs when price is reduced;

t_pk is a time of peak sales;

α and

γ

are curve fitting constants;

C is a residual sales rate at large values of t; and

N_c is a function representing a demand multiplier that is a function of price,

ii. predicting a maximum profit for the item of commerce by optimizing price or inventory or both over time in accordance with the aforesaid relationship and in accord with a relationship wherein profit is a function of sales revenues minus costs, and

B. using the optimized price or inventory or both in connection with making price decisions or inventory decisions or both with respect to marketing the item of commerce.

4. The method of claim 3 further comprising

from additional data that includes prices and unit sales of additional items, deriving additional relationships that represent variations of unit sales of the additional items as a function of price and time of sale,

predicting a maximum profit by optimizing prices or inventories or both over time accordance with the relationships, and

using the optimized prices or inventories or both over time in connection with making price decisions or inventory decisions or both with respect to the marketing of the items of commerce.

5. The method of claim 3 in which the item comprises a fashion item.

6. The method of claim 3 in which the item comprises a basic item.

7. The method of claim 3 in which the data comprises historical data.

8. The method of claim 3 in which the optimizing comprises applying a genetic algorithm optimization technique.

9. The method of claim 3 further comprising

applying a Monte Carlo simulation to the optimized price or inventory or both over time to generate a set of predicted outcomes and probability information for respective outcomes, and

using the result of the simulation to evaluate confidence in different ones of the outcomes.

10. The method of claim 3 in which the optimizing takes account of cross-correlation between items.

11. The method of claim 3 in which the relationship that represents variation of sales of the item as a function of price over time comprises an equation of a particular form.

12. The method of claim 3 in which the relationship that represents variation of sales of the item as a function of price over time comprises a trend line factor multiplied by a seasonal variation factor.

13. The method of claim 3 in which the relationship that represents variation of sales of the item as a function of price over time comprises a decaying exponential term.

14. The method of claim 3 in which the relationship that represents variation of sales of the item as a function of price over time comprises a demand multiplier term that expresses the relationship of demand for the item to a declining price.

16. The method of claim 3 in which the relationship that represents variation of sales of the item as a function of price over time comprises a decaying exponential term.

17. The method of claim 3, further comprising performing curve-fitting to determine one or more of the aforesaid constants.

18. The method of claim 17, wherein the curve-fitting step includes performing curve-fitting using a non-linear least-squares fit.

19. The method of claim 18, wherein the curve-fitting step includes performing the Levenberg-Marquart method.

20. The method of claim 3, further comprising performing curve-fitting to determine the demand multiplier N_c.

21. The method of claim 20, wherein the curve-fitting step includes fitting a curve to a table of markdown demand multipliers of sales increases versus mark-down percentages determine from raw sales data.

22. The method of claim 20, wherein the curve-fitting step includes determining N_c is a function of the relation

N_c=e^(m/j)β

·

e<sup>−

((t−

tmd)/hmd)κ</sup>

where β

is a constant, m is a markdown fraction, j is a constant, t_md is a time of markdown, t is time, h_md is a constant that depends on merchandise types and brand, and κ

is a constant.

15. The method of 14 in which the demand multiplier term comprises a decay factor which reflects a declining impact of a price drop with time, and an exponential factor based on the size of the price drop.

23. A machine-based method for making price decisions or inventory decisions or both with respect to the marketing of an item of commerce, the method comprising

A. executing, on a machine, software that configures the machine to perform the following steps

i. from data that includes prices and unit sales of the item of commerce for a succession of time periods during which the item was sold, deriving a relationship of the form

S(t)=N_o*N_c*t^γ

*[e<sup>−

(t−

tpk)α</sup>

+C]

 

where S(t) represents gross sales revenues of the item;

t is time;

N_o is a curve fitting parameter;

N_c is a demand multiplier representing an increase in sales that occurs when price is reduced;

t_pk is a time of peak sales;

α and

γ

are curve fitting constants;

C is a residual sales rate at large values of t; and

N_c is a function representing a demand multiplier that is a function of price,

ii. performing curve-fitting using a non-linear least-squares fit to determine one or more of the aforesaid constants,

iii. determining N_c is a function of the relation

N_c=e^(m/j)β

·

e<sup>−

((t−

tmd)/hmd)κ</sup>

 

where β

is a constant, m is a markdown fraction, j is a constant, t_{md is a time of markdown, t is time, hmd} is a constant that depends on merchandise types and brand, and κ

is a constant,

iv. predicting a maximum profit for the item of commerce by optimizing price or inventory or both over time in accordance with the aforesaid relationship and in accord with a relationship wherein profit is a function of sales revenues minus costs, and

B. using the optimized price or inventory or both in connection with making price decisions or inventory decisions or both with respect to marketing the item of commerce.

24. The method of claim 23, wherein the optimizing step includes applying a genetic algorithm optimization technique.