US-12626277-B2 - Systems and methods for continuous incrementality monitoring of tracked electronic communications

Abstract

A computer system for channel incrementality, including tracking, using first data associated with at least one user device, at least one characteristic of a user, and tracking, using second data associated with the at least one user device, at least one first interaction with at least two tracked communications. The system further generates weight data for at least two channels which weighs an impact that each channel has on the user's tendency to complete a transaction. The system receives information on a completed transaction and apportions the transaction based on the weight data.

Inventors

• Yan Rui Huang
• Scott Thomas Sheinbaum

Assignees

• COUPANG CORP.

Dates

Publication Date

20260512

Application Date

20231228

Claims (20)

1. 1 . A computer-implemented system for incrementality monitoring, the system comprising: a memory storing instructions; and at least one processor configured to execute the instructions to: store at least one parameter for testing one of at least two channels; determine the effectiveness of the one channel by testing the one channel based on the stored parameter; update a user-specific model which weighs the impact that each of the at least two channels has on a user's purchasing tendencies based on determining a weight allocated to the one channel falls outside a maximum or minimum value indicated by the determined effectiveness; track, via a tracker associated with one user device, one user interaction with a tracked communication leading up to a transaction, wherein the tracker is coded to detect one of a click on the tracked communication, swipe on the tracked communication, impression on the tracked communication, or mouse hovering over the tracked communication; apportion the transaction by feeding the one user interaction into the updated model; and retrieve a new keyword for the one channel based on the apportionment; display a new communication in response to receiving the new keyword on the one user device.
2. 2 . The system of claim 1 , wherein the at least one parameter comprises at least one of: parameters that indicate a period of time advertising will be turned on and a period of time that advertising will be turned off, parameters that indicate a period of time advertising will not be reduced and a period of time that advertising will be reduced, parameters that indicate a period of time advertising spending will not be reduced and a period of time that advertising spending will be reduced, a parameter that indicates a portion of advertising that will be turned off, a parameter that indicates a portion of advertising that will remain on, a parameter that indicates a portion of advertising spending that will be removed, a parameter that indicates a portion of advertising spending that will remain, a parameter that indicates a portion of a plurality of users that will receive advertising that is not reduced and a portion of the plurality of users that will receive reduced advertising, a parameter that indicates a portion of a plurality of users that will receive advertising and a portion of the plurality of users that will not receive advertising, or a first parameter for one of the at least two channels and a second parameter for another of the at least two channels.
3. 3 . The system of claim 2 , wherein the at least one parameter indicates a period of time that advertising will not be reduced and a period of time that advertising will be reduced; and wherein the at least one processor is further configured to: receive purchase data indicative of purchases made in the period of time that advertising is not reduced; and receive purchase data indicative of purchases made in the period of time that advertising is reduced.
4. 4 . The system of claim 3 , wherein the at least one processor is further configured to: determine the effectiveness of the at least one channel by comparing the purchase data indicative of purchases made in the period of time that advertising is not reduced to the purchase data indicative of purchases made in the period of time that advertising is reduced.
5. 5 . The system of claim 2 , wherein the parameter indicates a portion of a plurality of users that will receive advertising that is not reduced and a portion of the plurality of users that will receive reduced advertising; wherein the one processor is further configured to: receive purchase data indicative of purchases made by the portion of the plurality of users that receive advertising that is not reduced; and receive purchase data indicative of purchases made by the portion of the plurality of users that receive reduced advertising.
6. 6 . The system of claim 5 , wherein the at least one processor is further configured to: determine the effectiveness of the at least one channel by comparing the purchase data indicative of purchases made by the portion of the plurality of users that receive advertising that is not reduced to the purchase data indicative of purchases made by the portion of the plurality of users that receive reduced advertising.
7. 7 . The system of claim 6 , wherein updating the model comprises adjusting the weight allocated to the at least one channel based on the determined effectiveness.
8. 8 . The system of claim 7 , wherein the model is a logistic regression model and adjusting the weight of the one channel based on the determined effectiveness comprises setting a weight range for the at least one channel and adjusting the model to weigh the at least one channel based on the weight range.
9. 9 . The system of claim 1 , wherein the at least one processor is further configured to: determine the effectiveness of a second channel of the at least two channels; and update the model based on the determined effectiveness of the second channel.
10. 10 . The system of claim 1 , wherein the at least one processor is further configured to: repeatedly perform the determination of the effectiveness of the one channel after an increment of time; and repeatedly update the model based on the determined effectiveness after the increment of time.
11. 11 . A computer-implemented method for incrementality monitoring, comprising: storing at least one parameter for testing one of at least two channels; determining the effectiveness of the one channel by testing the one channel based on the stored parameter; updating a user-specific model which weighs the impact that each of the at least two channels has on a user's purchasing tendencies based on determining a weight allocated to the one channel falls outside a maximum or minimum value indicated by the determined effectiveness; tracking, via a tracker associated with one user device, one user interaction with a tracked communication leading up to a transaction, wherein the tracker is coded to detect one of a click on the tracked communication, swipe on the tracked communication, impression on the tracked communication, or mouse hovering over the tracked communication; apportioning the transaction by feeding the one user interaction into the updated model; and retrieving a new keyword for the one channel based on the apportionment; displaying a new communication in response to receiving the new keyword on the one user device.
12. 12 . The method of claim 11 , wherein the one parameter comprises one of: parameters that indicate a period of time advertising will be turned on and a period of time that advertising will be turned off, parameters that indicate a period of time advertising will not be reduced and a period of time that advertising will be reduced, parameters that indicate a period of time advertising spending will not be reduced and a period of time that advertising spending will be reduced, a parameter that indicates a portion of advertising that will be turned off, a parameter that indicates a portion of advertising that will remain on, a parameter that indicates a portion of advertising spending that will be removed, a parameter that indicates a portion of advertising spending that will remain, a parameter that indicates a portion of a plurality of users that will receive advertising that is not reduced and a portion of the plurality of users that will receive reduced advertising, a parameter that indicates a portion of a plurality of users that will receive advertising and a portion of the plurality of users that will not receive advertising, or a first parameter for one of the at least two channels and a second parameter for another of the at least two channels.
13. 13 . The method of claim 12 , wherein the at least one parameter indicates a period of time that advertising will not be reduced and a period of time that advertising will be reduced; and wherein method further comprises: receiving purchase data indicative of purchases made in the period of time that advertising is not reduced; and receiving purchase data indicative of purchases made in the period of time that advertising is reduced.
14. 14 . The method of claim 13 , further comprising: determining the effectiveness of the at least one channel by comparing the purchase data indicative of purchases made in the period of time that advertising is not reduced to the purchase data indicative of purchases made in the period of time that advertising is reduced.
15. 15 . The method of claim 12 , wherein the parameter indicates a portion of a plurality of users that will receive advertising that is not reduced and a portion of the plurality of users that will receive reduced advertising; wherein the method further comprises: receiving purchase data indicative of purchases made by the portion of the plurality of users that receive advertising that is not reduced; and receiving purchase data indicative of purchases made by the portion of the plurality of users that receive reduced advertising.
16. 16 . The method of claim 15 , wherein the method further comprises: determining the effectiveness of the at least one channel by comparing the purchase data indicative of purchases made by the portion of the plurality of users that receive advertising that is not reduced to the purchase data indicative of purchases made by the portion of the plurality of users that receive reduced advertising.
17. 17 . The method of claim 16 , wherein updating the model comprises adjusting the weight allocated to the at least one channel based on the determined effectiveness.
18. 18 . The method of claim 17 , wherein the model is a logistic regression model and adjusting the weight of the at least one channel based on the determined effectiveness comprises setting a weight range for the one channel and adjusting the model to weigh the one channel based on the weight range.
19. 19 . The method of claim 11 , further comprising: determining the effectiveness of a second channel of the at least two channels; and updating the model based on the determined effectiveness of the second channel.
20. 20 . A computer-implemented system for incrementality monitoring, the system comprising: a memory storing instructions; and at least one processor configured to execute the instructions to: store at least one parameter for testing at least one of at least two channels; determine the effectiveness of the at least one channel by testing the at least one channel based on the stored parameter; update a user-specific model which weighs the impact that each of the at least two channels has on a user's purchasing tendencies based on determining a weight allocated to the one channel falls outside a maximum or minimum value indicated by the determined effectiveness; wherein the model is a logistic regression model and updating the model comprises adjusting the weight allocated to the one channel based on the maximum or minimum value indicated by the determined effectiveness; track, via a tracker associated with one user device, at least one user interaction with a tracked communication leading up to a transaction, wherein the tracker is coded to detect at least one of a click on the tracked communication, swipe on the tracked communication, impression on the tracked communication, or mouse hovering over the tracked communication; apportion the transaction by feeding the at least one user interaction into the updated model; and retrieve a new keyword for the one channel based on the apportionment; display a new communication in response to receiving the new keyword on the at least one user device.

Description

TECHNICAL FIELD The present disclosure generally relates to computerized systems and methods for targeted electronic advertising. In particular, embodiments of the present disclosure relate to inventive and unconventional systems to apportion sales between different advertising channels, determine effectiveness of an advertising campaign, and perform monitoring of the advertising channels to ensure the models used in these processes are accurate. BACKGROUND Consumers react to advertisements on different channels in different manners. A consumer may be strongly impacted by advertisements they receive on a first channel and are significantly more likely to make a purchase after receiving an advertisement on that first channel. However, the same consumer may be less impacted by advertisements they receive on a second channel and are not as likely to make a purchase after receiving an advertisement on the second channel. Further, a consumer's decision to make a purchase is not solely a result of the advertisements they receive. Instead, a consumer may have certain baseline purchasing tendencies that must be removed when determining an impact of an advertisement on a consumer. Existing systems are not able to determine what impact user interactions with advertisements on an individual channel had on a consumer's purchase or whether the interactions had any impact at all. Therefore, there is a need for improved methods and systems for determining what impact individual advertising channels had on a consumer's purchase decision so companies can adjust marketing strategies accordingly. Further, in determining this impact, there is a need to remove a consumer's baseline purchasing tendencies. As described above, advertisements on a first advertising channel may have a different impact on a consumer than advertisements on a second advertising channel. Further, across consumers, certain types of advertisements may perform better on some advertising channels than others. A company may want to target certain types of advertisements on advertising channels where they are most effective. Existing systems are not able to determine what impact user interactions with advertisements on an individual channel had on a consumer's purchase or whether the interactions had any impact at all. Further, existing systems are not able to make changes to advertising on an individual channel based on the impact. Therefore, there is a need for improved methods and systems for determining what impact a type of advertisement has on an individual marketing channel so companies can adjust marketing strategies accordingly. Further, there is a need to make automatic marketing changes so that certain types of advertisements are on advertising channels where they are most effective. As described above, advertisements on a first advertising channel may have a different impact on a consumer than advertisements on a second advertising channel. Further, the impact of each of these advertising channels may change with time. The impact of an advertising channel needs to be updated to accurately determine the impact of an advertisement received on that advertising channel. Existing systems are not able to determine what impact user interactions with advertisements on an individual channel had on a consumer's purchase or whether the interactions had any impact at all. Further, existing systems are static and are not able to determine whether credit attributed to an advertising channel is still accurate over time. Therefore, there is a need for improved methods and systems to continuously monitoring changes to the impact an advertising channel has on consumers. Additionally, there is a need to update a model for apportioning a purchase based on the changes. Further, there is a need to apportion a consumer's purchase decision between multiple advertising channels based on the updated model. SUMMARY One aspect of the present disclosure is directed to a computer-implemented system comprising a memory storing instructions, and at least one processor configured to: track, using first data associated with at least one user device, at least one characteristic of a user, the user being associated with the at least one user device, track, using second data associated with the at least one user device, at least one first interaction with at least two tracked communications, each tracked communication being associated with a separate channel of at least two channels, generate weight data for the at least two channels which weighs an impact that each channel has on the user's tendency to complete a transaction. Wherein the generating weight data comprises: feeding the at least one characteristic of a user into a first model; feeding the at least one first interaction into the first model; and utilizing the first model to correlate a transaction completion tendency with the at least one characteristic and the at least one interaction. Further, the processor is configur