Sourcing goods based on pre-feature analytics

Abstract

In general, embodiments of the present invention provide systems, methods and computer readable media for collecting pre-feature data for a promotion, performing analytics on the pre-feature data, predicting the promotion sales velocity based on applying an odds model to the pre-feature data, and determining a quantity of goods to source based at least in part on the promotion sales velocity.

1 - 30 . (canceled) 
     
     
         31 . A system, comprising one or more computers and one or more storage devices storing instructions that are operable, when executed by the one or more computers, to cause the one or more computers to:
 train a machine learning odds model based on historical promotion activation data until tuning criteria for the machine learning odds model is satisfied, wherein the historical promotion activation data comprises a historical number of promotion activations over one or more promotion durations;   select a subset of consumer devices from a set of consumer devices based on first attributes of a promotion and second attributes of respective consumer identifiers associated with the subset of consumer devices;   distribute the promotion to the subset of consumer devices for rendering via respective electronic interfaces of the subset of consumer devices during a pre-feature promotion duration;   receive pre-feature data describing respective pre-feature impressions of the promotion distributed to subset of consumer devices, wherein the pre-feature data comprises response data collected from the respective pre-feature impressions and pre-feature promotion activation data resulting from the respective pre-feature impressions of the promotion, and wherein the pre-feature promotion activation data comprises a number of promotion activations in response to the respective pre-feature impressions of the promotion distributed to subset of consumer devices;   generate a predicted value for the promotion by applying at least a portion of the pre-feature data to the machine learning odds model, wherein the predicted value is a predicted number of promotion activations over a promotion duration that is different than the pre-feature promotion duration, and wherein the machine learning odds model comprises a scaling parameter that corrects for quantity of the pre-feature promotion activations based on a quantity of the pre-feature promotion activation data and a total quantity of goods being offered via the respective pre-feature impressions during the pre-feature promotion duration; and   determine inventory data for one or more other promotions based on the predicted value for the promotion.   
     
     
         32 . The system of  claim 31 , wherein the one or more storage devices store instructions that are operable, when executed by the one or more computers, to further cause the one or more computers to:
 select, from a set of predefined display positions, a display position for the promotion for rendering via the respective electronic interfaces based on one or more of the first attributes of the promotion.   
     
     
         33 . The system of  claim 31 , wherein the one or more storage devices store instructions that are operable, when executed by the one or more computers, to further cause the one or more computers to:
 select, from a set of predefined display positions, a display position for the promotion for rendering via the respective electronic interfaces based on one or more of the second attributes of the respective consumer identifiers.   
     
     
         34 . The system of  claim 31 , wherein the one or more storage devices store instructions that are operable, when executed by the one or more computers, to further cause the one or more computers to:
 configure a display size for the promotion for rendering via the respective electronic interfaces based on one or more of the first attributes of the promotion.   
     
     
         35 . The system of  claim 31 , wherein the one or more storage devices store instructions that are operable, when executed by the one or more computers, to further cause the one or more computers to:
 configure a display size for the promotion for rendering via the respective electronic interfaces based on one or more of the second attributes of the respective consumer identifiers.   
     
     
         36 . The system of  claim 31 , wherein the first attributes comprises promotion price features associated with a promotion price for the promotion, and wherein the one or more storage devices store instructions that are operable, when executed by the one or more computers, to further cause the one or more computers to:
 select the subset of consumer devices from a set of consumer devices based on the promotion price features.   
     
     
         37 . The system of  claim 31 , wherein the second attributes comprises location features associated with respective locations for respective consumer devices, and wherein the one or more storage devices store instructions that are operable, when executed by the one or more computers, to further cause the one or more computers to:
 select the subset of consumer devices from a set of consumer devices based on the location features.   
     
     
         38 . A computer-implemented method, comprising:
 training, by a computing device comprising a processor, a machine learning odds model based on historical promotion activation data until tuning criteria for the machine learning odds model is satisfied, wherein the historical promotion activation data comprises a historical number of promotion activations over one or more promotion durations;   selecting, by the computing device, a subset of consumer devices from a set of consumer devices based on first attributes of a promotion and second attributes of respective consumer identifiers associated with the subset of consumer devices;   distributing, by the computing device, the promotion to the subset of consumer devices for rendering via respective electronic interfaces of the subset of consumer devices during a pre-feature promotion duration;   receiving, by the computing device, pre-feature data describing respective pre-feature impressions of the promotion distributed to subset of consumer devices, wherein the pre-feature data comprises response data collected from the respective pre-feature impressions and pre-feature promotion activation data resulting from the respective pre-feature impressions of the promotion, and wherein the pre-feature promotion activation data comprises a number of promotion activations in response to the respective pre-feature impressions of the promotion distributed to subset of consumer devices;   generating, by the computing device, a predicted value for the promotion by applying at least a portion of the pre-feature data to the machine learning odds model, wherein the predicted value is a predicted number of promotion activations over a promotion duration that is different than the pre-feature promotion duration, and wherein the machine learning odds model comprises a scaling parameter that corrects for quantity of the pre-feature promotion activations based on a quantity of the pre-feature promotion activation data and a total quantity of goods being offered via the respective pre-feature impressions during the pre-feature promotion duration; and   determining, by the computing device, inventory data for one or more other promotions based on the predicted value for the promotion.   
     
     
         39 . The computer-implemented method of  claim 38 , further comprising:
 selecting, by the computing device and from a set of predefined display positions, a display position for the promotion for rendering via the respective electronic interfaces based on one or more of the first attributes of the promotion.   
     
     
         40 . The computer-implemented method of  claim 38 , further comprising:
 selecting, by the computing device and from a set of predefined display positions, a display position for the promotion for rendering via the respective electronic interfaces based on one or more of the second attributes of the respective consumer identifiers.   
     
     
         41 . The computer-implemented method of  claim 38 , further comprising:
 configuring, by the computing device, a display size for the promotion for rendering via the respective electronic interfaces based on one or more of the first attributes of the promotion.   
     
     
         42 . The computer-implemented method of  claim 38 , further comprising:
 configuring, by the computing device, a display size for the promotion for rendering via the respective electronic interfaces based on one or more of the second attributes of the respective consumer identifiers.   
     
     
         43 . The computer-implemented method of  claim 38 , wherein the first attributes comprises promotion price features associated with a promotion price for the promotion, and the selecting the subset of consumer devices from the set of consumer devices comprising selecting the subset of consumer devices from the set of consumer devices based on the promotion price features. 
     
     
         44 . The computer-implemented method of  claim 38 , wherein the second attributes comprises location features associated with respective locations for respective consumer devices, and the selecting the subset of consumer devices from the set of consumer devices comprising selecting the subset of consumer devices from the set of consumer devices based on the location features. 
     
     
         45 . A computer program product, stored on a computer readable medium, comprising instructions that when executed by one or more computers cause the one or more computers to:
 train a machine learning odds model based on historical promotion activation data until tuning criteria for the machine learning odds model is satisfied, wherein the historical promotion activation data comprises a historical number of promotion activations over one or more promotion durations;   select a subset of consumer devices from a set of consumer devices based on first attributes of a promotion and second attributes of respective consumer identifiers associated with the subset of consumer devices;   distribute the promotion to the subset of consumer devices for rendering via respective electronic interfaces of the subset of consumer devices during a pre-feature promotion duration;   receive pre-feature data describing respective pre-feature impressions of the promotion distributed to subset of consumer devices, wherein the pre-feature data comprises response data collected from the respective pre-feature impressions and pre-feature promotion activation data resulting from the respective pre-feature impressions of the promotion, and wherein the pre-feature promotion activation data comprises a number of promotion activations in response to the respective pre-feature impressions of the promotion distributed to subset of consumer devices;   generate a predicted value for the promotion by applying at least a portion of the pre-feature data to the machine learning odds model, wherein the predicted value is a predicted number of promotion activations over a promotion duration that is different than the pre-feature promotion duration, and wherein the machine learning odds model comprises a scaling parameter that corrects for quantity of the pre-feature promotion activations based on a quantity of the pre-feature promotion activation data and a total quantity of goods being offered via the respective pre-feature impressions during the pre-feature promotion duration; and   determine inventory data for one or more other promotions based on the predicted value for the promotion.   
     
     
         46 . The computer program product of  claim 45 , further comprising instructions that when executed by the one or more computers cause the one or more computers to:
 select, from a set of predefined display positions, a display position for the promotion for rendering via the respective electronic interfaces based on one or more of the first attributes of the promotion.   
     
     
         47 . The computer program product of  claim 45 , further comprising instructions that when executed by the one or more computers cause the one or more computers to:
 select, from a set of predefined display positions, a display position for the promotion for rendering via the respective electronic interfaces based on one or more of the second attributes of the respective consumer identifiers.   
     
     
         48 . The computer program product of  claim 45 , further comprising instructions that when executed by the one or more computers cause the one or more computers to:
 configure a display size for the promotion for rendering via the respective electronic interfaces based on one or more of the first attributes of the promotion.   
     
     
         49 . The computer program product of  claim 45 , further comprising instructions that when executed by the one or more computers cause the one or more computers to:
 configure a display size for the promotion for rendering via the respective electronic interfaces based on one or more of the second attributes of the respective consumer identifiers.   
     
     
         50 . The computer program product of  claim 45 , wherein the first attributes comprises promotion price features associated with a promotion price for the promotion, and the computer program product further comprising instructions that when executed by the one or more computers cause the one or more computers to:
 select the subset of consumer devices from a set of consumer devices based on the promotion price features.