Method, apparatus, and computer program product for forecasting demand

• US 9,947,022 B1
• Filed: 03/14/2013
• Issued: 04/17/2018
• Est. Priority Date: 10/04/2012
• Status: Active Grant

• Current Assignee: Groupon Incorporated
• Original Assignee: Groupon Incorporated
• Inventors: Shariff, Shafiq, Parekh, Rajesh, Beaty, Bob, Hou, Feili
• Primary Examiner(s): Gills, Kurtis
• Application Number:
  US13/826,333
• Time in Patent Office: 1,860 Days
• Field of Search: 705 731
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 20/10 using kernel methods, e.g. support vector machines [SVM]
• CPC Class Codes: G06N 5/02 Knowledge representation; Symbolic representation
• CPC Class Codes: G06N 5/04 Inference or reasoning models
• CPC Class Codes: G06Q 10/06311 Scheduling, planning or task assignment for a person or group
• CPC Class Codes: G06Q 10/06315 Needs-based resource requirements planning or analysis
• CPC Class Codes: G06Q 10/0635 Risk analysis of enterprise or organisation activities
• CPC Class Codes: G06Q 10/0639 Performance analysis of employees; Performance analysis of enterprise or organisation operations
• CPC Class Codes: G06Q 10/06393 Score-carding, benchmarking or key performance indicator [KPI] analysis
• CPC Class Codes: G06Q 10/067 Enterprise or organisation modelling
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data
• CPC Class Codes: G06Q 30/0202 Market predictions or forecasting for commercial activities
• CPC Class Codes: G06Q 30/0282 Rating or review of business operators or products
• CPC Class Codes: G06Q 30/0609 Buyer or seller confidence or verification
• CPC Class Codes: G06Q 40/00 Finance; Insurance; Tax strategies; Processing of corporate or income taxes

Abstract

Provided herein are systems, methods and computer readable media for forecasting demand. An example method comprises generating a virtual offer for one or more combinations of a category or sub-category, location, and price range, accessing consumer data comprising one or more users and user data related to each of the one or more users, calculating a probability that a particular user would buy a particular offer in a particular time frame for at least a portion of the plurality of users and for each of the virtual offers, and determining an estimated number of units of to be sold for at least a portion of the one or more virtual offers as a function of at least the probability associated with each of the one or more virtual offers.

Claims

1. A method for virtual offer demand forecasting comprising:

generating, by one or more processors via a virtual offer generation module, a virtual offer for each combination of a plurality of attributes, wherein the attributes comprise only price data, one product category or one service category, and location data,

wherein the generation of the virtual offer for each combination of the plurality of attributes comprises;

accessing both a category taxonomy model and a location taxonomy model to identify a plurality of sub-category and hyper-location pairs, the category taxonomy model defining a hierarchical structure of service categories and sub-categories that is found in one or more local or hyper-local regions and the location taxonomy model defining a hierarchical structure that is based on locations, sub-locations, and hyper-local regions, and

utilizing a set of guidelines to match at least one of one or more price ranges to the sub-category and hyper-location pair resulting in a set of virtual offers comprised of each combination of a plurality of attributes;

accessing, via a consumer data database, consumer data comprising one or more users and user data related to each of the one or more users;

calculating, by one or more processors via a probability generation module, a probability value that a particular user from the one or more users would buy a particular offer from the set of virtual offers comprised of each combination of a plurality of attributes in a particular time frame for at least a portion of the one or more users and for each of the plurality of the virtual offers;

generating a pivot table using the list of consumers and their associated probabilities for each of the plurality of virtual offers;

converting the pivot table into a listing of the plurality of virtual offers and related probabilities for each consumer;

determining an estimated number of units to be sold for at least a portion of the plurality of virtual offers as a function of at least the probability associated with each of the plurality of virtual offers;

correlating the at least one of the plurality virtual offers to a real promotion by matching the price data, product category or service category, and the location data of the at least one of the plurality of virtual offers to price data, a product or service category and the location data of the real promotion;

accessing information indicative of the available inventory of the real promotion;

calculating a residual demand as a function of the forecasted demand and the available inventory, and

dynamically adjusting the residual demand, subsequent to a process in which merchants are contacted based on their availability to service the residual demand, based on a change in the available inventory of the real promotion due to a closed contract and diversity constraint information requiring merchant diversity and diversity among categories, sub-categories, locations, and price ranges.

2. The method of claim 1, further comprising calculating a forecasted demand by

adjusting the forecasted demand as a function of one or more external factors.

3. The method of claim 1, further comprising:

generating a ranked user list for each of the one or more users comprising each of the one or more virtual offers and an associated calculated probability;

assigning a voting value to each of the one or more virtual offers on each ranked user list; and

generating a single list ranking each of the one or more virtual offers by adding the voting value of each of the one or more virtual offers received on each ranked user list.

4. The method of claim 3, wherein the selecting of the one or more virtual offers further comprises:

calculating a projected revenue value for each virtual offer based on prior performance data per unit and the estimated number of units to be sold;

ranking the one or more virtual offers by the projected revenue; and

selecting one or more virtual offers to include in the forecasted demand based on the ranking of each of the one or more virtual offers by revenue and a pre-selected portfolio mix.

5. The method of claim 2, wherein the one or more external factors includes at least a seasonality component.

6. The method according to claim 3, wherein the voting value is derived from a position of the one or more virtual offers on the ranked lists generated for each of the one or more users, wherein at least a predetermined number of virtual offers on each ranked list are assigned a first voting value.

7. The method according to claim 1, wherein one or more combinations of category or sub-category, location and price range are generated for a specified time period.

22. The method according to claim 1, wherein generation of the virtual offer for one or more combinations of a category, sub-category, location, and price ranges comprises:

accessing a taxonomy model and virtual offer builder, the taxonomy model defining a hierarchical structure of categories and sub-categories that may be found in one or more local or hyper-local regions, a hierarchical structure that is based on target consumer categories, or a hierarchical structure that is based on service; and

generating virtual offers according to a set of guidelines, the guideline operable to provide limitations on prices or price ranges that one or more sub-sets of categories or sub-categories are paired within a virtual offer, provide limitations according to prior performance of the virtual offers at each of one or more prices or price ranges and prior performance of certain categories, or provide limitations according to which categories or sub-categories are available during particular time periods,

wherein the virtual offers are not offered for sale by a seller nor offered to a consumer.

23. The method according to claim 1, wherein the probability value is only generated for a subset of virtual offers within a predetermined distance of the particular user of the one or more users.

24. The method according to claim 3, wherein the voting value is derived from a position of the one or more virtual offers on the ranked lists generated for each of the one or more users, wherein a first voting value is assigned to a first virtual offer, the first virtual offer having a highest probability value, and a second voting value is assigned to a second virtual offer, the second virtual offer having a probability value less than the highest probability value, and the second voting value being a predetermined percentage less than the first voting value.

8. An apparatus comprising at least one processor and at least one memory including computer program code, the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus at least to:

generate, via a virtual offer generation module, a virtual offer for each combination of a plurality of attributes, wherein the attributes comprise only price data, one product category or one service category, and location data,

wherein the generation of the virtual offer for each combination of the plurality of attributes comprises;

accessing both a category taxonomy model and a location taxonomy model to identify a plurality of sub-category and hyper-location pairs, the category taxonomy model defining a hierarchical structure of service categories and sub-categories that is found in one or more local or hyper-local regions and the location taxonomy model defining a hierarchical structure that is based on locations, sub-locations, and hyper-local regions resulting in a set of virtual offers comprised of each combination of a plurality of attributes, and

utilize a set of guidelines to match at least one of one or more price ranges to the sub-category and hyper-location pair;

access, via a consumer data database, consumer data comprising one or more users and user data related to each of the one or more users;

calculate, via a probability generation module, a probability value that a particular user from the one or more users would buy a particular offer from the set of virtual offers comprised of each combination of a plurality of attributes in a particular time frame for at least a portion of the one or more users and for each of the plurality of the virtual offers;

generate a pivot table using the list of consumers and their associated probabilities for each of the plurality of virtual offers;

convert the pivot table into a listing of the plurality of virtual offers and related probabilities for each consumer;

determining an estimated number of units to be sold for at least a portion of the plurality of the virtual offers as a function of at least the probability associated with each of the plurality of virtual offers;

correlate the at least one of the plurality virtual offers to a real promotion by matching the price data, product category or service category, and the location data of the at least one of the plurality of virtual offers to price data, a product or service category and the location data of the real promotion;

access information indicative of the available inventory of the real promotion;

calculate a residual demand as a function of the forecasted demand and the available inventory; and

dynamically adjust the residual demand, subsequent to a process in which merchants are contacted based on their availability to service the residual demand, based on a change in the available inventory of the real promotion due to a closed contract and diversity constraint information requiring merchant diversity and diversity among categories, sub-categories, locations, and price ranges.

9. The apparatus of claim 8, wherein the at least one memory and computer program code are configured to, with the at least one processor, cause the apparatus to calculate a forecasted demand by

adjusting the forecasted demand as a function of one or more external factors.

10. The apparatus of claim 8, wherein the at least one memory and computer program code are configured to, with the at least one processor, cause the apparatus to:

generate a ranked user list for each of the one or more users comprising each of the one or more virtual offers and an associated calculated probability;

assign a voting value to each of the one or more virtual offers on each ranked user list; and

generate a single list ranking each of the one or more virtual offers by adding the voting value of each of the one or more virtual offers received on each ranked user list.

11. The apparatus of claim 10, wherein the selecting of the one or more virtual offers further comprises:

calculating a projected revenue value for each virtual offer based on prior performance data per unit and the estimated number of units to be sold;

ranking the one or more virtual offers by the projected revenue; and

selecting one or more virtual offers to include in the forecasted demand based on the ranking of each of the one or more virtual offers by revenue and a pre-selected portfolio mix.

12. The apparatus of claim 9, wherein the one or more external factors includes at least a seasonality component.

13. The apparatus of claim 10, wherein the voting value is derived from a position of the one or more virtual offers on the ranked lists generated for each of the one or more users, wherein at least a predetermined number of virtual offers on each ranked list are assigned a first voting value.

14. The apparatus of claim 8, wherein one or more combinations of category or sub-category, location and price range are generated for a specified time period.

15. A computer program product comprising at least one non-transitory computer-readable storage medium having computer-executable program code instructions stored therein, the computer-executable program code instructions comprising program code instructions for:

generating, via a virtual offer generation module, a virtual offer for each combination of a plurality of attributes, wherein the attributes comprise only, price data, one product category or one service category, and location data,

wherein the generation of the virtual offer or each combination of the plurality of attributes comprises;

accessing both a category taxonomy model and a location taxonomy model to identify a plurality of sub-category and hyper-location pairs, a sub-category and hyper-location pair, the category taxonomy model defining a hierarchical structure of service categories and sub-categories that is found in one or more local or hyper-local regions and the location taxonomy model defining a hierarchical structure that is based on locations, sub-locations, and hyper-local regions, and

utilizing a set of guidelines to match at least one of one or more price ranges to the sub-category and hyper-location pair resulting in a set of virtual offers comprised of each combination of a plurality of attributes;

accessing, via a consumer data database, consumer data comprising one or more users and user data related to each of the one or more users;

calculating, via a probability generation module, a probability value that a particular user from the one or more users would buy a particular offer from the set of virtual offers comprised of each combination of a plurality of attributes in a particular time frame for at least a portion of the one or more users and for each of the virtual offers;

generating a pivot table using the list of consumers and their associated probabilities for each of the plurality of virtual offers;

converting the pivot table into a listing of the plurality of virtual offers and related probabilities for each consumer;

determining an estimated number of units to be sold for at least a portion of the one or more virtual offers as a function of at least the probability associated with each of the one or more virtual offers;

correlating the at least one of the plurality virtual offers to a real promotion by matching the price data, product category or service category, and the location data of the at least one of the plurality of virtual offers to price data, a product or service category and the location data of the real promotion;

accessing information indicative of the available inventory of the real promotion;

calculating a residual demand as a function of the forecasted demand and the available inventory and

dynamically adjusting the residual demand, subsequent to a process in which merchants are contacted based on their availability to service the residual demand, based on a change in the available inventory of the real promotion due to a closed contract and diversity constraint information requiring merchant diversity and diversity among categories, sub-categories, locations, and price ranges.

16. A computer program product according to claim 15, wherein the computer-executable program code portions further comprise program code instructions for calculating a forecasted demand by

adjusting the forecasted demand as a function of one or more external factors.

17. A computer program product according to claim 15, wherein the computer-executable program code portions further comprise program code instructions for:

generating a ranked user list for each of the one or more users comprising each of the one or more virtual offers and an associated calculated probability;

assigning a voting value to each of the one or more virtual offers on each ranked user list; and

generating a single list ranking each of the one or more virtual offers by adding the voting value of each of the one or more virtual offers received on each ranked user list.

18. A computer program product according to claim 17, wherein the selecting of the one or more virtual offers further comprises:

calculating a projected revenue value for each virtual offer based on prior performance data per unit and the estimated number of units to be sold;

ranking the one or more virtual offers by the projected revenue; and

selecting one or more virtual offers to include in the forecasted demand based on the ranking of each of the one or more virtual offers by revenue and a pre-selected portfolio mix.

19. A computer program product according to claim 16, wherein the one or more external factors includes at least a seasonality component.

20. A computer program product according to claim 17, wherein the voting value is derived from a position of the one or more virtual offers on the ranked lists generated for each of the one or more users, wherein at least a predetermined number of virtual offers on each ranked list are assigned a first voting value.

21. A computer program product according to claim 15, wherein one or more combinations of category or sub-category, location and price range are generated for a specified time period.