Methods and systems for predicting consumer behavior from transaction card purchases

• US 10,430,803 B2
• Filed: 12/23/2008
• Issued: 10/01/2019
• Est. Priority Date: 12/23/2008
• Status: Active Grant

• Current Assignee: Mastercard International Incorporated (MasterCard Incorporated)
• Original Assignee: Mastercard International Incorporated (MasterCard Incorporated)
• Inventors: Bene, Marc Del, Hu, Po, Nambiar, Anant, Salazar, Daniel G.
• Primary Examiner(s): Dickerson, Tiphany B
• Application Number:
  US12/342,925
• Publication Number:
  US 20100161379A1
• Time in Patent Office: 3,934 Days
• Field of Search: 705 731
• CPC Class Codes: G06Q 10/06375 Prediction of business process outcome or impact based on a proposed change
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0202 Market predictions or forecasting for commercial activities
• CPC Class Codes: G06Q 30/0204 Market segmentation

Abstract

A computer-based method for predicting consumer behavior is provided. The method is performed using a computer system coupled to a database. The method includes recording consumer data in the database for each consumer of a global population of consumers including historical purchases made by each consumer using a transaction card, defining a life event by assigning spending variables to the life event, determining a sample group of consumers that are experiencing the life event based on the consumer data stored within the database with respect to the spending variables, generating a predictive model based on historical purchases made by consumers within the sample group, and applying the predictive model to predict each consumer within the global population that will experience the life event. The predictive model is applied using the computer system. A list of consumers predicted to experience the life event within a predetermined time period is output.

Claims

1. A computer-based method for predicting consumer behavior within a predetermined time period, said method performed using a payment processor computer device coupled to a database, said method comprising:

recording, by the payment processor computer device, electronic consumer data in the database for each consumer of a global population of consumers including historical purchases made by each consumer, including at least one electronic transaction initiated by a consumer using a transaction card at a point-of-sale (POS) device;

defining, by the payment processor computer device, a life event by assigning spending variables to the life event, wherein a spending variable represents a quantity of consumer spending associated with one or more of (i) a particular merchant and (ii) a particular type of good or service;

generating an anticipated spend for each spending variable assigned to the life event by extrapolating a historical actual spend based on the historical purchases, wherein the anticipated spend represents anticipated spending in each spending variable assigned to the life event;

calculating a residual value of a determined variance between the anticipated spend and an actual spend for each consumer by calculating a quotient of (i) the difference between the anticipated spend and the actual spend and (ii) the anticipated spend;

determining that a residual value of a determined variance between the anticipated spend and the actual spend for each consumer exceeds a predefined threshold;

assigning, by the payment processor computer device, the consumer to a sample group, wherein the sample group represents consumers that are experiencing the life event;

generating, by the payment processor computer device, a predictive model based on historical purchases made by consumers within the sample group;

applying, by the payment processor computer device, the predictive model to predict each consumer within the global population and outside of the sample group that will experience the life event; and

outputting, by the payment processor computer device, a list of consumers outside of the sample group that are predicted to experience the life event within the predetermined time period.

2. A computer-based method in accordance with claim 1 wherein performing residual analysis further comprises:

analyzing historical purchases for each consumer within the global population having an account in the database during a historical time period for each spending variable in a consumption bundle of the life event;

determining a historical actual spending curve for each consumer within the global population during the historical time period using the historical purchases made in the consumption bundle;

determining an anticipated spending curve for each consumer within the global population during a second time period after the historical time period using the historical actual spending curve, wherein the anticipated spending curve is determined before the second time period occurs;

collecting purchase data for each consumer within the global population over the second time period; and

determining an actual spending curve during the second time period for each consumer within the global population using the collected purchase data, wherein the actual spending curve is determined after the second time period occurs.

3. A computer-based method in accordance with claim 2 wherein performing residual analysis further comprises:

comparing the actual spending curve and the anticipated spending curve for each consumer within the global population during the second time period;

determining a variance between the actual spending curve and the anticipated spending curve for each consumer within the global population; and

assigning consumers within the global population to the sample group using the residual value of the determined variance.

4. A computer-based method in accordance with claim 1 further comprising:

causing a customized offer to be sent to the at least one consumer on the output list for a product related to a life event being experienced by the at least one consumer, based on a the spending variable assigned to the life event.

5. A computer-based method in accordance with claim 1 further comprising refining the predictive model using actual spending by at least one consumer on the output list of consumers based on purchases made on a transaction card.

6. A computer-based method in accordance with claim 1 wherein recording consumer data in the database for each consumer of a global population further comprises:

recording transaction card purchase data for each consumer within the global population to the database; and

recording third party data about each consumer within the global population to the database.

7. A computer-based method in accordance with claim 1 wherein defining a life event by assigning spending variables to the life event further comprises:

selecting a life event a consumer may experience; and

assigning at least one spending variable to the selected life event to form a consumption bundle for the selected life event.

8. A computer-based method in accordance with claim 7 wherein defining a life event by assigning spending variables to the life event further comprises assigning at least one demographic variable to the life event, the consumption bundle comprising the at least one assigned demographic variable and the at least one assigned spending variable.

9. A computer-based method in accordance with claim 1 wherein generating a predictive model based on historical purchases made by consumers within the sample group further comprises:

analyzing each consumer in the sample group during a historical time period to determine spending trends that are common to consumers within the sample group based on the historical purchases, wherein a spending trend represents at least one of an increase, a decrease or no change in the amount of spending in a spending variable; and

generating a predictive model that includes the determined spending trends.

10. A computer-based method in accordance with claim 1 wherein applying the predictive model to predict each consumer within the global population that will experience the life event further comprises:

applying the predictive model to all consumers within the global population having an account in the database;

modeling actual spending trends of each consumer within the global population using the predictive model, the predictive model including a set of spending trends; and

when the actual spending trends of a consumer within the global population are similar to the set of spending trends included within the predictive model, predicting that the consumer will experience the life event.

11. A computer-based method in accordance with claim 1 wherein defining a life event further comprises defining the life event as one of having a new baby, sending a child to college, and purchasing a new home.

12. A computer for predicting behavior of a consumer within a predetermined time period based on the consumer'"'"'s purchases made using a transaction card, said computer comprising a processor, computer-readable instructions executable by the processor, and a database, said computer configured to:

record electronic consumer data in said database for each consumer of a global population of consumers including historical purchases made by each consumer, including at least one electronic transaction initiated by a consumer using a transaction card at a point-of-sale (POS) device;

define a life event by assigning spending variables to the life event, wherein a spending variable represents a quantity of consumer spending associated with one or more of (i) a particular merchant and (ii) a particular type of good or service;

generate an anticipated spend by extrapolating a historical actual spend based on the historical purchases, wherein the anticipated spend represents anticipated spending in each spending variable assigned to the life event;

calculate a residual value of a determined variance between the anticipated spend and an actual spend for each consumer by calculating a quotient of (i) the difference between the anticipated spend and the actual spend and (ii) the anticipated spend;

determine that a residual value of a determined variance between the anticipated spend and the actual spend for each consumer exceeds a predefined threshold;

assign the consumer to a sample group, wherein the sample group represents consumers that are experiencing the life event;

generate a predictive model based on historical purchases made by consumers within the sample group, wherein the predictive model is generated using the processor;

apply the predictive model to predict each consumer within the global population and outside of the sample group that will experience the life event; and

output a list of consumers outside of the sample group that are predicted to experience the life event within the predetermined time period.

13. A computer in accordance with claim 12 further configured to:

analyze historical purchases of a consumer within the global population having an account in said database during a historical time period for each spending variable in a consumption bundle of the life event;

determine a historical actual spending curve during the historical time period for the consumer using the historical purchases made in the consumption bundle;

determine an anticipated spending curve during a second time period after the historical time period for the consumer using the historical actual spending curve, wherein the anticipated spending curve is determined before the second time period occurs;

collect purchase data for the consumer over the second time period;

determine an actual spending curve during the second time period for the consumer using the collected purchase data, wherein the actual spending curve is determined after the second time period occurs;

compare the actual spending curve and the anticipated spending curve for the consumer during the second time period;

determine a variance between the actual spending curve and the anticipated spending curve; and

assign the consumer to the sample group using the residual value of the determined variance.

14. A computer in accordance with claim 12 further configured to:

map anticipated consumer needs to at least one offer stored within the database based on the defined life event; and

provide the least one offer to the consumer.

15. A computer in accordance with claim 12 further configured to refine the predictive model using actual spending of the consumers within the sample group based on purchases made on transaction cards by at least one of adding a spending variable to the definition of the life event and removing a spending variable from the definition of the life event.

16. A computer in accordance with claim 12 further configured to:

select a life event a consumer may experience; and

assign at least one spending variable to the selected life event to form a consumption bundle for the selected life event.

17. A computer in accordance with claim 12 further configured to:

analyze each consumer in the sample group during a historical time period to determine spending trends that are common to consumers within the sample group based on the recorded consumer data, wherein a spending trend represents at least one of an increase, a decrease or no change in the amount of spending in a spending variable; and

generate a predictive model that includes the determined spending trends.

18. A computer in accordance with claim 17 further configured to:

apply the predictive model to each consumer within the global population;

model actual spending trends of each consumer within the global population with the predictive model; and

when actual spending trends of a consumer within the global population are similar to spending trends within the predictive model, predict that the consumer will experience the life event.

19. A network based system for predicting behavior of a consumer within a predetermined time period based on the consumer'"'"'s purchases made using a transaction card, said system comprising:

a client computing device;

a database for storing information; and

a server computing device comprising a processor and computer-readable instructions executable by said processor, said server computing device configured to be coupled to said client computing device and said database, said server computing system further configured to;

record electronic consumer data in said centralized database for each consumer of a global population of consumers including historical purchases made by each consumer, including at least one electronic transaction initiated by a consumer using a transaction card at a point-of-sale (POS) device;

define a life event by assigning spending variables to the life event, wherein a spending variable represents a quantity of consumer spending associated with one or more of (i) a particular merchant and (ii) a particular type of good or service;

generate an anticipated spend by extrapolating a historical actual spend based on the historical purchases, wherein the anticipated spend represents anticipated spending in each spending variable assigned to the life event;

calculate a residual value of a determined variance between the anticipated spend and an actual spend for each consumer by calculating a quotient of (i) the difference between the anticipated spend and the actual spend and (ii) the anticipated spend;

determine that a residual value of a determined variance between the anticipated spend and the actual spend for each consumer exceeds a predefined threshold;

assign the consumer to a sample group, wherein the sample group represents consumers that are experiencing the life event;

generate a predictive model based on historical purchases made by consumers within the sample group;

apply the predictive model to predict each consumer within the global population and outside of the sample group that will experience the life event; and

output a list of consumers outside of the sample group that are predicted to experience the life event within the predetermined time period.

20. A network based system in accordance with claim 19, wherein said server computing device is further configured to:

analyze historical purchases of a consumer within the global population having an account in said database during a historical time period for each spending variable in a consumption bundle of the life event;

determine a historical actual spending curve during the historical time period for the consumer using the historical purchases made in the consumption bundle;

determine an anticipated spending curve during a second time period after the historical time period for the consumer using the historical actual spending curve, wherein the anticipated spending curve is determined before the second time period occurs;

collect purchase data for the consumer over the second time period;

determine an actual spending curve during the second time period for the consumer using the collected purchase data, wherein the actual spending curve is determined after the second time period occurs;

compare the actual spending curve and the anticipated spending curve for the consumer during the second time period;

determine a variance between the actual spending curve and the anticipated spending curve; and

assign the consumer to the sample group using the residual value of the determined variance.

21. A network based system in accordance with claim 19, wherein said server computing device is further configured to:

analyze each consumer in the sample group during a historical time period to determine spending trends that are common to consumers within the sample group, wherein a spending trend represents at least one of an increase, a decrease or no change in the amount of spending in a spending variable;

generate a predictive model that includes the determined spending trends;

apply the predictive model to each consumer within the global population;

model actual spending trends of each consumer within the global population with the spending trends included within the predictive model; and

when actual spending trends of a consumer within the global population are similar to spending trends within the predictive model, predict that the consumer will experience the life event.

22. A computer program embodied on a non-transitory computer readable storage medium for predicting consumer behavior, said program comprising at least one code segment that, when executed by a computing device including at least one processor in communication with a memory, causes the computing device to:

record electronic consumer data in the database for each consumer of a global population of consumers including historical purchases made by each consumer, including at least one electronic transaction initiated by a consumer using a transaction card at a point-of-sale (POS) device;

define a life event by assigning spending variables to the life event, wherein a spending variable represents a quantity of consumer spending associated with one or more of (i) a particular merchant and (ii) a particular type of good or service;

generate an anticipated spend by extrapolating a historical actual spend based on the historical purchases, wherein the anticipated spend represents anticipated spending in each spending variable assigned to the life event;

calculate a residual value of a determined variance between the anticipated spend and an actual spend for each consumer by calculating a quotient of (i) the difference between the anticipated spend and the actual spend and (ii) the anticipated spend;

determine that a residual value of a determined variance between the anticipated spend and the actual spend for each consumer exceeds a predefined threshold;

assign the consumer to a sample group, wherein the sample group represents consumers that are experiencing the life event;

generate a predictive model based on historical purchases made by consumers within the sample group;

apply the predictive model to predict each consumer within the global population and outside of the sample group that will experience the life event; and

output a list of consumers outside of the sample group that are predicted to experience the life event within the predetermined time period.