System and method for phased estimation and correction of promotion effects

Patent metadata

Field	Value
Publication number	US-2016307218-A1
Application number	US-201514686896-A
Country	US
Kind code	A1
Filing date	Apr 15, 2015
Priority date	Apr 15, 2015
Publication date	Oct 20, 2016
Grant date	—

Abstract

Systems, methods, and other embodiments are disclosed that are configured to generate promotion effects for use by a demand forecast model. In one embodiment, a first regression analysis is performed on historical performance data for an item to generate first promotion effect values for a plurality of promotion components. The first promotion effect values are compared to a plurality of rules. If none of the rules are violated by the first promotion effect values, the first promotion effect values are output as final promotion effect values. If at least one of the rules is violated by the first promotion effect values, a phased operation is performed on the historical performance data. The phased operation operates on subsets of the promotion components of the historical performance data over at least two regression analysis phases to generate second promotion effect values.

First claim

What is claimed is:

1 . A method implemented by a computing device configured to execute a computer application, wherein the computer application is configured to process data in electronic form, the method comprising:
performing a first regression analysis on historical performance data for an item to generate first promotion effect values for a plurality of promotion components associated with the item, wherein the historical performance data includes at least unit sales data and data associated with the plurality of promotion components across a plurality of time periods; comparing the first promotion effect values to a plurality of first rules to determine if one or more of the plurality of first rules is violated by one or more of the first promotion effect values; if none of the plurality of first rules are violated, outputting the first promotion effect values to an output data structure as final promotion effect values; and if at least one of the plurality of first rules is violated, performing a phased operation on the historical performance data, wherein the phased operation operates on subsets of the plurality of promotion components of the historical performance data over at least two regression analysis phases to generate second promotion effect values.

2 . The method of claim 1 , further comprising:
comparing the second promotion effect values to the plurality of first rules; and outputting values of the second promotion effect values, that do not violate the plurality of first rules, to the output data structure as the final promotion effect values.

3 . The method of claim 1 , further comprising:
comparing the second promotion effect values to a plurality of second rules; and outputting values of the second promotion effect values, that do not violate the a plurality of second rules, to the output data structure as the final promotion effect values.

4 . The method of claim 1 , wherein the data associated with the plurality of promotion components is obtained by reading, from a data structure, data that includes, for each time period of a plurality of time periods, a price discount value and a plurality of Boolean values indicating that individual promotion components of the plurality of promotion components were either active or inactive for a corresponding time period.

5 . The method of claim 1 , wherein generating the first promotion effect values includes generating at least one promotion effect value that is a price elasticity value representing a measure of how a demand for an item responds to a change in a price of the item.

6 . The method of claim 1 , wherein generating the second promotion effect values includes generating at least one promotion effect value that is a price elasticity value representing a measure of how a demand for an item responds to a change in a price of the item.

7 . The method of claim 1 , wherein the plurality of promotion components include a price discount component and at least a television advertisement component, a radio advertisement component, a newspaper advertisement component, an internet advertisement component, an email advertisement component, or an in-store advertisement component.

8 . The method of claim 1 , wherein the performing the phased operation on the historical performance data includes:
performing a second regression analysis on a first subset of the plurality of promotion components to generate a price elasticity value of the second promotion effect values, wherein the first subset of the plurality of promotion components represents price discounting; generating de-priced performance data by removing price discount effects from the historical performance data based at least in part on the price elasticity value; and performing a third regression analysis on a second subset of the plurality of promotion components, based at least in part on the de-priced performance data, to generate at least one remaining value of the second promotion effect values.

9 . The method of claim 8 , wherein the second subset of the plurality of promotion components includes at least one of a television advertisement component, a radio advertisement component, a newspaper advertisement component, an internet advertisement component, am email advertisement component, and an in-store advertisement component.

10 . The method of claim 8 , wherein the at least one remaining value includes at least one of a television advertisement value, a radio advertisement value, a newspaper advertisement value, an internet advertisement value, an email advertisement value, or an in-store advertisement value.

11 . A computing system, comprising:
a regression module, including instructions stored in a non-transitory computer-readable medium, configured to:
(i) generate first promotion effect values for a plurality of promotion components associated with an item by performing a first regression analysis on historical performance data for the item, wherein the historical performance data includes at least unit sales data and data associated with the plurality of promotion components across a plurality of time periods, and
(ii) generate second promotion effect values by performing a phased operation on the historical performance data, wherein the phased operation operates on subsets of the plurality of promotion components of the historical performance data over at least two regression analysis phases; and
a comparator module, including instructions stored in the non-transitory computer-readable medium, configured to:
(i) compare the first promotion effect values to a plurality of rules to determine if one or more of the plurality of rules is violated by one or more of the first promotion effect values,
(ii) compare the second promotion effect values to the plurality of rules to determine if one or more of the plurality of rules is violated by one or more of the second promotion effect values, and
(iii) control the regression module to perform the phased operation when one or more of the plurality of rules is violated by one or more of the first promotion effect values.

12 . The computing system of claim 11 , further comprising a rules module, including instructions stored in the non-transitory computer-readable medium, configured to store the plurality of rules in at least one data structure, wherein the at least one data structure is accessible to the comparator module.

13 . The computing system of claim 11 , further comprising a database device configured to store at least the historical performance data.

14 . The computing system of claim 11 , further comprising a visual user interface module, including instructions stored in the non-transitory computer-readable medium, configured to:
provide a graphical user interface; and facilitate inputting of the historical performance data for the item into the regression module.

15 . The computing system of claim 14 , further comprising a rules module, including instructions stored in the non-transitory computer-readable medium, configured to operably interact with the visual user interface module via the graphical user interface to facilitate generation of the plurality of rules.

16 . The computing system of claim 14 , further comprising a display screen configured to display and facilitate user interaction with at least the graphical user interface.

17 . A non-transitory computer-readable medium storing computer-executable instructions that are part of an algorithm that, when executed by a computer, cause the computer to perform functions, wherein the inst

Assignees

• Oracle Int Corp

Inventors

• Lei Ming
• Popescu Catalin

Classifications

• G06Q30/0242

  Determining effectiveness of advertisements · CPC title

• G06Q30/0202 Primary

  Market predictions or forecasting for commercial activities · CPC title

• G06Q30/0211

  Determining the effectiveness of discounts or incentives · CPC title