Apparatus and method for enhanced message targeting

Patent metadata

Field	Value
Publication number	US-10990987-B2
Application number	US-201916552531-A
Country	US
Kind code	B2
Filing date	Aug 27, 2019
Priority date	Mar 30, 2015
Publication date	Apr 27, 2021
Grant date	Apr 27, 2021

Abstract

A method, apparatus, and computer program product are disclosed for improved machine learning using a statistical model. In the context of an apparatus, some example embodiments include a processor configured to cause retrieval of information regarding a plurality of consumers, and modeling circuitry configured to train a statistical model of the plurality of consumers based on the retrieved information, and predict, using the statistical model, an incremental booking value associated with the promotion for each consumer of the plurality of consumers. The processor is further configured to select a subset of the plurality of consumers for receiving impressions of the promotion. Some example embodiments may further include communications circuitry configured to transmit an impression of the promotion to each consumer in the subset of the plurality of consumers.

First claim

What is claimed is:

1. An apparatus for improved machine learning using a statistical model, the apparatus comprising at least one processor and at least one non-transitory computer-readable memory, having computer-coded instructions thereon, that in execution with the at least one processor configure the apparatus to:
train the statistical model based on retrieved information regarding a plurality of consumers, and
predict, using the statistical model, an incremental booking value associated with a promotion for each consumer of the plurality of consumers by at least:
estimating, using the statistical model, a first expected revenue wherein the first expected revenue is estimated based on a first set of input information comprising at least a first promotion indicator indicating the consumer would have access to the promotion for a first time; period; and
calculatimg the incremental booking value based on a second expected revenue and the first expected revenue, wherein the second expected revenue is based on a second set of input information comprising at least a second promotion indicator indicating the consumer would not have access to the promotion; and

select a subset of the plurality of consumers for whom the predicted incremental booking value satisfies a predefined threshold; and
transmit an impression of the promotion to each consumer in the subset of the plurality of consumers.

2. The apparatus of claim 1 , is the apparatus further configured to cause retrieval of information regarding the plurality of consumers, wherein to cause retrieval of the information the apparatus is configured to:
receive electronic marketing information from one or more of the plurality of consumers, from one or more merchants, from a memory, or from a combination thereof.

3. The apparatus of claim 2 , wherein the electronic marketing information includes at least one of: a number of purchases within a past thirty days; a number of purchases made by a consumer through a promotion and marketing service; a number of impressions received via a consumer's mobile device that have been clicked by the consumer; a total revenue that a consumer has provided over all promotions; consumer tenure information; a path that a consumer used to sign up for a promotion and marketing service; a number of impressions received via a consumer's regular email that have been clicked on by the consumer;
an indication of whether a consumer has received an impression of a particular promotion; a number of promotions purchased by a consumer within three zip codes of a given location within a prior thirty days; an average cost of each of a consumer's bookings; a number of active channel subscriptions by a consumer; a total number of bookings that a consumer has made; a highest price of any promotion purchased by a consumer via a promotion and marketing service; a number of promotions between $0 and $15 that have been shown to a consumer; and a median price of a consumer's bookings.

4. The apparatus of claim 1 , wherein the apparatus is configured to predict an expected incremental booking value of the promotion for each consumer by:
estimating, using the statistical model, a first expected booking value that would be received from the consumer during a first time period in an instance in which the consumer has access to the promotion;
estimating, using the statistical model, a second expected booking value that would be received from the consumer in an instance in which the consumer does not have access to the promotion; and
calculating a first difference value by subtracting the second expected booking value from the first expected booking value.

5. The apparatus of claim 4 , wherein the expected incremental booking value comprises the first difference value.

6. The apparatus of claim 4 , to the apparatus further configured to:
calculate a second difference value by subtracting a discount value associated with the promotion from the first difference value,
wherein the expected incremental booking value comprises the second difference value.

7. The apparatus of claim 1 wherein the apparatus is configured to select the subset of the plurality of consumers for receiving impressions of the promotion by:
ranking the plurality of consumers by incremental booking values associated with the promotion that correspond to each of the plurality of consumers; and
selecting a predetermined percentage of highest ranked consumers,
wherein the subset of the plurality of consumers to target comprises the predetermined percentage of highest ranked consumers.

8. A method for improved machine learning using a statistical model executed by a processor, the method comprising:
training a statistical model based on retrieved information regarding a plurality of consumers;
predicting, using the statistical model, an incremental booking value associated with a promotion for each consumer of the plurality of consumers by at least:
estimating, using the statistical model, a first expected revenue, wherein the first expected revenue is estimated based on a first set of input information comprising at least a first promotion indicator indicating the consumer would have access to the promotion for a first time period; and
calculating the incremental booking value based on a second expected revenue and the first expected revenue, wherein the second expected revenue is based on a second set of input information comprising at least a second promotion indicator indicating the consumer would not have access to the promotion,

selecting a subset of the plurality of consumers for whom the predicted incremental booking value satisfies a predefined threshold; and
transmitting an impression of the promotion to each consumer in the subset of the plurality of consumers.

9. The method of claim 8 , wherein retrieving information regarding the plurality of consumers includes:
receiving electronic marketing information from one or more of the plurality of consumers, from one or more merchants, from a memory, or from a combination thereof.

10. The method of claim 9 , wherein the electronic marketing information includes at least one of: a number of purchases within a past thirty days; a number of purchases made by a consumer through a promotion and marketing service; a number of impressions received via a consumer's mobile device that have been clicked by the consumer; a total revenue that a consumer has provided over all promotions; consumer tenure information; a path that a consumer used to sign up for a promotion and marketing service; a number of impressions received via a consumer's regular email that have been clicked on by the consumer;
an indication of whether a consumer has received an impression of a particular promotion; a number of promotions purchased by a consumer within three zip codes of a given location within a prior thirty days; an average cost of each of a consumer's bookings; a number of active channel subscriptions by a consumer; a total number of bookings that a consumer has made; a highest price of any promotion purchased by a consumer via a promotion and marketing service; a number of promotions between $0 and $15 that have been shown to a consumer; and a median price of a consumer's bookings.

11. The method of claim 8 , wherein predicting an expected incremental booking value of the promotion for each consumer includes:
estimating, by the modeling circuitry and using the statistical model, a first expected booking value that would be received from the consumer during a first time period in an instance in which the consumer has access to the promotion;
estimating, by the modeling circuitry and using the statistical model, a second e

Assignees

• Groupon Inc

Inventors

• Parekh Rajesh Girish
• Balakrishnan Raju
• Zhang Guoxian
• Gaurav Prashant

Classifications

• G06N20/20

  Ensemble learning · CPC title

• G06Q30/0211

  Determining the effectiveness of discounts or incentives · CPC title

• G06N20/00

  Machine learning · CPC title

• G06Q30/0201 Primary

  Market modelling; Market analysis; Collecting market data · CPC title