Systems and methods for generating and recommending promotions in a design matrix

• US 10,706,438 B2
• Filed: 08/24/2016
• Issued: 07/07/2020
• Est. Priority Date: 03/13/2013
• Status: Active Grant

• Current Assignee: Maplebear Inc.
• Original Assignee: Eversight, Inc. (Maplebear Inc.)
• Inventors: Montero, Michael, Eldredge, Jamie, Brockardt, John, Hsieh, Kevin, Murphy, Lynn, Solotaroff, Jacob, Moran, David
• Primary Examiner(s): Andrei, Radu
• Application Number:
  US15/246,348
• Publication Number:
  US 20170053305A1
• Time in Patent Office: 1,413 Days
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0239 Online discounts or incentives
• CPC Class Codes: G06Q 30/0244 Optimization
• CPC Class Codes: G06Q 30/0269 based on user profile or attribute
• CPC Class Codes: G06Q 30/0277 Online advertisement

Abstract

Systems and methods for the generation of a promotional design matrix with recommendation overlay are provided. A product, discount, quantity and offer structure selection are received from a user, or are imputed based upon previous offers. Variants for the discount vale and quantity are generated by incrementing these values. The quantity and its variants are each assigned to a column of a design matrix, while the discount and variants are assigned to the design matrix rows. At each intersection of a row and column of the design matrix a plurality of offer structures based upon the offer structure selection is generated. The system applies recommendation logic to flag offers that are most likely to be successful based upon their performance, feasibility and diversity. This flagging may include placing an icon, such as a star, next to the recommended offers.

Claims

1. A computer implemented method for generating a design matrix with recommendations that is performed by a promotion optimization system that includes a processor and memory, the method comprising:

receiving a product, a discount, a quantity and at least one offer structure selection;

generating at least two variants for the discount by altering one increment up and down from the discount by a first set percentage change from the discount when the discount is at or below a threshold and by a second set percentage change that is larger than the first set percentage change from the discount when the discount above the threshold;

generating at least two variants for the quantity by increment up and down from the quantity by a set unit number, or by 90 and 110 percent of the quantity;

assigning the quantity and the at least two quantity variants to columns of a design matrix;

assigning the discount and the at least two discount variants to rows of a design matrix;

calculating at each intersection of a row and column of the design matrix a plurality of offer structures based upon the offer structure selection;

predicting performance of each offer structure in the design matrix using a model derived from past offer experiments;

flagging a subset of offer structures in the design matrix with a visual cue; and

displaying the design matrix with flagged offer structures to a marketing user.

2. The method of claim 1, wherein the product, discount, quantity and at least one offer structure selection are received from a user.

3. The method of claim 1, wherein at least one of the product, discount, quantity and at least one offer structure selection are imputed using past offer performance.

4. The method of claim 1, wherein the discount is a percentage or a dollar reduction.

5. The method of claim 4, wherein at least two discount variants are a set percentage change from the discount.

6. The method of claim 5, wherein at least two discount variants are rounded to an even dollar amount after the percentage difference is calculated.

7. The method of claim 1, further comprising generating a recommendation of offers most likely to be successful and flagging the recommended offers.

8. The method of claim 7, wherein the recommendation is based on a score calculated for each offer in the design matrix that is based upon past offers.

9. The method of claim 8, wherein the past offers are scored for performance, diversity and feasibility.

10. The method of claim 9, wherein the past offers are scored by multiplying a first weight to performance, multiplying a second weight to diversity, multiplying a third weight to feasibility, and summing the resulting values.

11. The method of claim 9, wherein the offer performance is a measure of offer redemption, downloads, click-through, shares, or actual sales lift.

12. The method of claim 9, wherein the offer diversity is a measure of how different the offer is from other offers, and is computed by tabulating all offer variable values and inserting 1 into the table for variable values that are the same between the offers and 0 for different variable values.

13. The method of claim 9, wherein the offer feasibility is the frequency the offer has been previously implemented.

14. The method of claim 9, further comprising generating a variable value listing for the offers in the design matrix.

15. The method of claim 14, further comprising determining which of the variable values were present in each of the past offers.

16. The method of claim 15, further comprising determining degree of similarity between the offers in the design matrix and past offers by tabulating all offer variable values and inserting 1 into the table for variable values that are the same between the offers in the design matrix and the past offers, and 0 for different variable values.

17. The method of claim 16, wherein the past offer score is used to model the score for the offers in the design matrix based upon the degree of similarity.

18. The method of claim 8, wherein the recommendation is for offers in the design matrix above a threshold.

19. The method of claim 8, wherein the flagging includes displaying a visual icon next to the recommended offers.