System, method and computer program product for geo-specific vehicle pricing

• US 10,296,929 B2
• Filed: 06/30/2011
• Issued: 05/21/2019
• Est. Priority Date: 06/30/2011
• Status: Active Grant

• Current Assignee: TrueCar, Inc.
• Original Assignee: TrueCar, Inc.
• Inventors: Swinson, Michael, Dhillon, Param Pash Kaur, Sullivan, Thomas J., Liu, Xingchu
• Primary Examiner(s): Weisenfeld, Aryan E
• Application Number:
  US13/173,357
• Publication Number:
  US 20130006876A1
• Time in Patent Office: 2,882 Days
• Field of Search: 705306, 705 11, 705 80, 705300-318, 705329, 705342-348, 705500
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0205 Location or geographical consideration
• CPC Class Codes: G06Q 30/0206 Price or cost determination based on market factors

Abstract

Disclosed are embodiments for the aggregation and analysis of vehicle prices via a geo-specific model. Data may be collected at various geo-specific levels such as a ZIP-Code level to provide greater data resolution. Data sets taken into account may include demarcation point data sets and data sets based on vehicle transactions. A demarcation point data set may be based on consumer market factors that influence car-buying behavior. Vehicle transactions may be classified into data sets for other vehicles having similar characteristics to the vehicle. A geo-specific statistical pricing model may then be applied to the data sets based on similar characteristics to a particular vehicle to produce a price estimation for the vehicle.

Claims

1. A method for pricing a vehicle in a geographic region, the method comprising:

obtaining, from a plurality of sources by a vehicle data system having a processor and a non-transitory computer readable medium, historical vehicle transaction data at varying levels of geographic resolution including at least a Zip Code level, the vehicle data system communicatively connected to the plurality of sources, user devices, and point of sale locations over a network;

storing, by the vehicle data system, the historical vehicle transaction data with vehicle features obtained from the plurality of sources in a data store, the vehicle features including year, make, model, and body (YMMB);

selecting, from the data store by the vehicle data system, a data set containing the historical vehicle transaction data and combinations of YMMB;

classifying, by the vehicle data system, the historical vehicle transaction data into a plurality of bins according to corresponding combinations of YMMB such that each of the plurality of bins includes one or more transactions for vehicles having the same combination of YMMB, and wherein all of the vehicle transaction data in the data set is classified into the plurality of bins;

subsequent to classifying the historical vehicle transaction data into the plurality of bins, determining, by the vehicle data system, a geographic resolution at which a number of transactions in every bin of the plurality of bins meets or exceeds a minimum threshold, the determining comprising determining the number of transactions in each of the plurality of bins beginning at the Zip Code level and relative to a geographic hierarchy having levels of geographic resolution which includes, in order of increasing coarseness, at least the Zip Code level, a Zip Code to Zip Code Tabulation Area (ZCTA) level, and a sub-Designated Market Areas (DMA) level, until the number of transactions in every bin of the plurality of bins meets or exceeds the minimum threshold, the geographic resolution thus determined by the vehicle data system representing a least coarse geographic resolution for the data set selected by the vehicle data system from the data store, the least coarse geographic resolution corresponding to a geographic region;

applying, by the vehicle data system, a geo-specific vehicle pricing model at the highest possible geographic resolution for the data set, wherein application of the geo-specific vehicle pricing model comprises;

determining a degradation factor value and a maximum transaction age over which transactions in the historical vehicle transaction data are not used in the geo-specific vehicle pricing model, the determining comprising analyzing historical performance of the historical vehicle transaction data using combinations of degradation factor values and transaction ages, choosing the combination of degradation factor value and transaction age which have the greatest relative performance, and weighting each transaction in the data set based on corresponding transaction age and the degradation factor value;

determining geo-specific socioeconomic data for a set of geo-specific socioeconomic variables to account for geographic differences in consumer behavior, the geo-specific socioeconomic data specific to the geographic region and including a number of point of sale locations for a specific combination of YMMB in the geographic region, and adjusting pricing for the geographic region based on the geo-specific socioeconomic data;

determining inventory data for a set of supply and demand variables, the inventory data including a number of days a vehicle of the specific combination of YMMB spent in a corresponding point of sale location before the vehicle is sold and a number of historical sales of vehicles having the specific combination of YMMB based on the transactions in the historical vehicle transaction data that meet or exceed the minimum transaction age and that thus are used in the geo-specific vehicle pricing model;

determining vehicle-specific features for a set of vehicle-specific variables, the vehicle-specific features including a vehicle body type; and

determining a predicted margin ratio representing a ratio of price over cost particular to a bin of the plurality of bins that corresponds to the specific combination of YMMB, the predicted margin ratio determined by the vehicle data system utilizing a regression model in which the set of geo-specific socioeconomic variables, the set of supply and demand variables, and the set of vehicle-specific variables are provided as weighted inputs so as to account for effects of the geo-specific socioeconomic data, the inventory data, and the vehicle-specific features on the predicted margin ratio;

generating, by the vehicle data system, a geo-specific price estimate for a vehicle of interest in the specific combination of YMMB within the geographic region, the generating utilizing the predicted margin ratio for the bin of the plurality of bins that corresponds to the specific combination of YMMB and a cost of the vehicle of interest; and

providing, by the vehicle data system, the geo-specific price estimate for the vehicle of interest to a website on the Internet for display on a client computing device over the network.

2. A method according to claim 1, further comprising:

adjusting the geo-specific price estimate to account for an incentive that is presently active for the vehicle of interest within the geographic region.

3. A method according to claim 1, further comprising:

assigning a temporal weight to each transaction of the transactions in the historical vehicle transaction data based on age.

4. A method according to claim 1, wherein the least coarse geographic resolution for the data set comprises a DMA region, a DMA group, a DMA, a subDMA, a ZCTA, or a Zip Code.

5. A method according to claim 1, wherein the geo-specific socioeconomic data further comprises a median home price in the geographic region as compared to a national median home price.

15. A method according to claim 1, further comprising:

terminating application of the geo-specific vehicle pricing model if the number of transactions in one or more bins of the plurality of bins does not meet or exceed the minimum threshold.

6. A vehicle data system for pricing a vehicle in a geographic region, the system comprising:

at least one processor communicatively connected to a network; and

at least one non-transitory computer readable storage medium storing instructions translatable by the at least one processor to perform the steps of;

obtaining, from a plurality of sources, historical vehicle transaction data at varying levels of geographic resolution including a Zip Code level, the vehicle data system communicatively connected to the plurality of sources, user devices, and point of sale locations over a network;

storing the historical vehicle transaction data with vehicle features obtained from the plurality of sources in a data store, the vehicle features including year, make, model, and body (YMMB);

selecting, from the data store, a data set containing the historical vehicle transaction data and combinations of YMMB;

classifying the historical vehicle transaction data into a plurality of bins according to corresponding combinations of YMMB such that each of the plurality of bins includes one or more transactions for vehicles having the same combination of YMMB, and wherein all of the vehicle transaction data in the data set is classified into the plurality of bins;

subsequent to classifying the historical vehicle transaction data into the plurality of bins, determining a geographic resolution at which a number of transactions in every bin of the plurality of bins meets or exceeds a minimum threshold, the determining comprising determining the number of transactions in each of the plurality of bins beginning at the Zip Code level and relative to a geographic hierarchy having levels of geographic resolution which includes, in order of increasing coarseness, at least the Zip Code level, a Zip Code to Zip Code Tabulation Area (ZCTA) level, and a sub-Designated Market Areas (DMA) level, until the number of transactions in every bin of the plurality of bins meets or exceeds the minimum threshold, the geographic resolution thus determined by the vehicle data system representing a least coarse geographic resolution for the data set selected by the vehicle data system from the data store, the highest possible geographic resolution corresponding to a geographic region;

applying a geo-specific vehicle pricing model at the highest possible geographic resolution for the data set, wherein application of the geo-specific vehicle pricing model comprises;

determining a degradation factor value and a maximum transaction age over which transactions in the historical vehicle transaction data are not used in the geo-specific vehicle pricing model, the determining comprising analyzing historical performance of the historical vehicle transaction data using combinations of degradation factor values and transaction ages, choosing the combination of degradation factor value and transaction age which have the greatest relative performance, and weighting each transaction in the data set based on corresponding transaction age and the degradation factor value;

determining geo-specific socioeconomic data for a set of geo-specific socioeconomic variables to account for geographic differences in consumer behavior, the geo-specific socioeconomic data specific to the geographic region and including a number of point of sale locations for a specific combination of YMMB in the geographic region, and adjusting pricing for the geographic region based on the geo-specific socioeconomic data;

determining inventory data for a set of supply and demand variables, the inventory data including a number of days a vehicle of the specific combination of YMMB spent in a corresponding point of sale location before the vehicle is sold and a number of historical sales of vehicles having the specific combination of YMMB based on the transactions in the historical vehicle transaction data that meet or exceed the minimum transaction age and that thus are used in the geo-specific vehicle pricing model;

determining vehicle-specific features for a set of vehicle-specific variables, the vehicle-specific features including a vehicle body type; and

determining a predicted margin ratio representing a ratio of price over cost particular to a bin of the plurality of bins that corresponds to the specific combination of YMMB, the predicted margin ratio determined by the vehicle data system utilizing a regression model in which the set of geo-specific socioeconomic variables, the set of supply and demand variables, and the set of vehicle-specific variables are provided as weighted inputs so as to account for effects of the geo-specific socioeconomic data, the inventory data, and the vehicle-specific features on the predicted margin ratio;

generating a geo-specific price estimate for a vehicle of interest within the locality and in the specific combination of YMMB within the geographic region, the generating utilizing the predicted margin ratio for the bin of the plurality of bins that corresponds to the specific combination of YMMB and a cost of the vehicle of interest; and

providing the geo-specific price estimate for the vehicle of interest to a website on the Internet for display on a client computing device over the network.

7. A system according to claim 6, wherein the instructions are further translatable by the at least one processor to perform:

adjusting the geo-specific price estimate to account for an incentive that is presently active for the vehicle of interest within the geographic region.

8. A system according to claim 6, wherein the instructions are further translatable by the at least one processor to perform:

assigning a temporal weight to each transaction of the transactions in the historical vehicle transaction data based on age.

9. A system according to claim 6, wherein the least coarse geographic resolution for the data set comprises a DMA region, a DMA group, a DMA, a subDMA, a ZCTA, or a Zip Code.

10. A system according to claim 6, wherein the geo-specific socioeconomic data further comprises a median home price in the geographic region as compared to a national median home price.

16. A system according to claim 6, wherein the instructions are further translatable by the at least one processor to perform:

terminating application of the geo-specific vehicle pricing model if the number of transactions in one or more bins of the plurality of bins does not meet or exceed the minimum threshold.

11. A computer program product for pricing a vehicle in a geographic region, the computer program product comprising at least one non-transitory computer readable medium storing instructions translatable by at least one processor of a vehicle data system to perform the steps of:

obtaining, from a plurality of sources, historical vehicle transaction data at varying levels of spatial resolution including a Zip Code level, the vehicle data system communicatively connected to the plurality of sources, user devices, and point of sale locations over a network;

storing the historical vehicle transaction data with vehicle features obtained from the plurality of sources in a data store, the vehicle features including year, make, model, and body (YMMB);

selecting, from the data store, a data set containing the historical vehicle transaction data and combinations of YMMB;

classifying the historical vehicle transaction data into a plurality of bins according to corresponding combinations of YMMB such that each of the plurality of bins includes one or more transactions for vehicles having the same combination of YMMB, and wherein all of the vehicle transaction data in the data set is classified into the plurality of bins;

subsequent to classifying the historical vehicle transaction data into the plurality of bins, determining a geographic resolution at which a number of transactions in every bin of the plurality of bins meets or exceeds a minimum threshold, the determining comprising determining the number of transactions in each of the plurality of bins beginning at the Zip Code level and relative to a geographic hierarchy having increasingly coarse levels of geographic resolution which includes, in order of increasing coarseness, at least the Zip Code level, a Zip Code to Zip Code Tabulation Area (ZCTA) level, and a sub-Designated Market Areas (DMA) level, until the number of transactions in every bin of the plurality of bins meets or exceeds the minimum threshold, the geographic resolution thus determined by the vehicle data system representing a least coarse geographic resolution for the data set selected by the vehicle data system from the data store, the least coarse geographic resolution corresponding to a geographic region;

applying a geo-specific price vehicle pricing model at the highest possible geographic resolution for the data set, wherein application of the geo-specific vehicle pricing model comprises;

determining a degradation factor value and a maximum transaction age over which transactions in the historical vehicle transaction data are not used in the geo-specific vehicle pricing model, the determining comprising analyzing historical performance of the historical vehicle transaction data using combinations of degradation factor values and transaction ages, choosing the combination of degradation factor value and transaction age which have the greatest relative performance, and weighting each transaction in the data set based on corresponding transaction age and the degradation factor value;

determining geo-specific socioeconomic data for a set of geo-specific socioeconomic variables to account for geographic differences in consumer behavior, the geo-specific socioeconomic data specific to the geographic region and including a number of point of sale locations for a specific combination of YMMB in the geographic region, and adjusting pricing for the geographic region based on the geo-specific socioeconomic data;

determining inventory data for a set of supply and demand variables, the inventory data including a number of days a vehicle of the specific combination of YMMB spent in a corresponding point of sale location before the vehicle is sold and a number of historical sales of vehicles having the specific combination of YMMB based on the transactions in the historical vehicle transaction data that meet or exceed the minimum transaction age and that thus are used in the geo-specific vehicle pricing model;

determining vehicle-specific features for a set of vehicle-specific variables, the vehicle-specific features including a vehicle body type; and

determining a predicted margin ratio representing a ratio of price over cost particular to a bin of the plurality of bins that corresponds to the specific combination of YMMB, the predicted margin ratio determined by the vehicle data system utilizing a regression model in which the set of geo-specific socioeconomic variables, the set of supply and demand variables, and the set of vehicle-specific variables are provided as weighted inputs so as to account for effects of the geo-specific socioeconomic data, the inventory data, and the vehicle-specific features on the predicted margin ratio;

generating a geo-specific price estimate for a vehicle of interest within the geographic region and in the bin of the plurality of bins that corresponds to the specific combination of YMMB, the generating utilizing the predicted margin ratio for the bin of the plurality of bins that corresponds to the specific combination of YMMB and a cost of the vehicle of interest; and

providing the geo-specific price estimate for the vehicle interest to a website on the Internet for display on a client computing device over the network.

12. A computer program product according to claim 11, wherein the instructions are further translatable by the at least one processor to perform:

adjusting the geo-specific price estimate to account for an incentive that is presently active for the vehicle of interest within the geographic region.

13. A computer program product according to claim 11, wherein the instructions are further translatable by the at least one processor to perform:

assigning a temporal weight to each transaction of the transactions in the historical vehicle transaction data based on age.

14. A computer program product according to claim 11, wherein the least coarse geographic resolution for the data set comprises a DMA region, a DMA group, a DMA, a subDMA, a ZCTA, or a Zip Code.

17. A computer program product according to claim 11, wherein the instructions are further translatable by the at least one processor to perform:

terminating application of the geo-specific vehicle pricing model if the number of transactions in one or more bins of the plurality of bins does not meet or exceed the minimum threshold.

18. A computer program product according to claim 11, wherein the geo-specific socioeconomic data further comprises a median home price in the geographic region as compared to a national median home price.