System and method for analysis and presentation of used vehicle pricing data

• US 10,108,989 B2
• Filed: 12/31/2013
• Issued: 10/23/2018
• Est. Priority Date: 07/28/2011
• Status: Active Grant

• Current Assignee: TrueCar, Inc.
• Original Assignee: TrueCar, Inc.
• Inventors: Swinson, Michael D., Laughlin, Isaac Lemon, Ramanuja, Meghashyam Grama, Semeniuk, Mikhail, Liu, Xingchu
• Primary Examiner(s): Yesildag, Mehmet
• Application Number:
  US14/145,252
• Publication Number:
  US 20140114726A1
• Time in Patent Office: 1,757 Days
• Field of Search: None
• CPC Class Codes: G06Q 10/06 Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0206 Price or cost determination based on market factors
• CPC Class Codes: G06Q 30/0278 Product appraisal

Abstract

Systems and methods for the aggregation, analysis, and display of data for used vehicles are disclosed. Historical transaction data for used vehicles may be obtained and processed to determine pricing data, where this determined pricing data may be associated with a particular configuration of a vehicle. The user can then be presented with an interface pertinent to the vehicle configuration utilizing the aggregated data set or the associated determined data where the user can make a variety of determinations. This interface may, for example, be configured to present the historical transaction data visually, with the pricing data such as a trade-in price, a list price, an expected sale price or range of sale prices, market low sale price, market average sale price, market high sale price, etc. presented relative to the historical transaction data.

Claims

1. A vehicle data system comprising:

a processor;

a non-transitory computer-readable medium; and

stored instructions executable by the processor to perform;

receiving or collecting, through application interfaces of the vehicle data system, data from distributed data sources, the data from the distributed data sources including historical sales transaction data comprising;

individual historical sales transaction data for a plurality of sold vehicles having a plurality of vehicle configurations sold in a plurality of regions,

vehicle residual value data for the plurality of vehicle configurations, and

geo-specific data for the plurality of vehicle configurations,

wherein each vehicle configuration in the plurality of vehicle configurations comprises a year, make and model, and

wherein the individual historical sales transaction data for the plurality of sold vehicles comprises sale prices and vehicle specific usage or maintenance data for the plurality of sold vehicles;

storing, in a data store, the historical sales transaction data, the vehicle residual value data, and the geo-specific data for the plurality of vehicle configurations;

performing a process divided into a back end process and a front end online process, the back end process performed at a time interval and asynchronously to the front end online process,

wherein the back end process comprises;

generating exponential decay estimate models for the plurality of vehicle configurations, each exponential decay estimate model modelling predicted residual values for a corresponding vehicle configuration for a given age input based on fitting an exponential decay curve to residual value data associated with the corresponding vehicle configuration;

generating residual value models for the plurality of vehicle configurations, each residual value model modelling residual value for a corresponding vehicle configuration at a transaction level as a function of predicted residual values for sold vehicles of the corresponding vehicle configuration determined by applying the exponential decay estimate model, thus generated in the back end process by the vehicle data system for the corresponding vehicle configuration, to individual historical sales transaction data for the corresponding vehicle configuration, regions associated with the individual historical sales transaction data for the corresponding vehicle configuration, and vehicle specific usage or maintenance data from the individual historical sales transaction data for sold vehicles of the corresponding configuration;

applying clustering rules to classify the historical sales transaction data for the plurality of sold vehicles into a plurality of bins based on vehicle characteristics and regions;

after classifying the historical sales transaction data for the plurality of sold vehicles into the plurality of bins, constructing research datasets for vehicle pricing, the research datasets comprising temporally-weighted historical data, goo-specific socioeconomic data, vehicle-specific attributes, vehicle usage or maintenance data and days-to-sell vehicle data;

constructing price ratio models based on a multivariable regression analysis, the price ratio models modelling a predicted price ratio for a vehicle i in a bin q of the plurality of bins as a function of a mean average price ratio for the bin q, a set of regression variables representing the vehicle-specific attributes, the vehicle usage or maintenance data, the geo-specific socioeconomic data, and the days-to-sell vehicle data, wherein a price ratio model is fitted to data in a dataset of the research datasets thus constructed in the back end process by the vehicle data system to find regression coefficients of the price ratio model;

storing, in the data store, the exponential decay estimate models, the residual value models, the research datasets, and the price ratio models generated in the back end process by the vehicle data system; and

wherein the front end online process comprises;

presenting a user interface on a client device over a network or Internet, the user interface comprising a web-based form with fields for entry of user input data through a client browser running on the client device;

receiving the user input data via the web-based form from the client device over the network or Internet through a network communication channel, the user input data comprising information about a used vehicle in a locale, including year, make, model, trim, mileage, and condition of the used vehicle;

receiving or obtaining socio-economic data specific to the locale from the user input data, the socio-economic data specific to the locale comprising income data, census data, home prices, or workforce participation data specific to the locale;

applying a set of rules to the locale and the year, make, model, trim, mileage, and condition of the used vehicle to determine a bin from the plurality of bins for the used vehicle;

constructing regression variables for the used vehicle based on the year, make, model, trim, mileage, and condition of the used vehicle and the socio-economic data specific to the locale;

determining an average price ratio of the determined bin;

applying the price ratio model using the average price ratio for the determined bin, the constructed regression variables for the used vehicle, and the regression coefficients of the price ratio model to determine a price ratio for the used vehicle;

selecting a residual value model for the used vehicle from the plurality of residual value models generated in the back end process by the vehicle data system based on the year, make, model, trim, mileage, and condition of the used vehicle;

determining a depreciated value of the used vehicle using the selected residual value model;

determining a final price for the used vehicle as a function of the depreciated value of the used vehicle and the price ratio for the used vehicle;

generating a responsive web page including the final price for the used vehicle and user interface controls linked to at least one of the regression variables for the used vehicle, the user interface controls configured to allow real-time adjustment of the linked regression variable in the responsive web page; and

communicating the responsive web page to the client device, wherein the responsive web page is generated and communicated to the client device in response the vehicle data system receiving the user input data from the client device.

2. The system of claim 1, wherein the front end online process further comprises:

receiving, through the user interface on the client device over the network or Internet, updated user input data comprising information about the used vehicle in the locale, including year, make, model, trim, mileage or condition;

adjusting the regression variables constructed for the used vehicle;

adjusting the final price determined for the used vehicle based on the updated user input data using the selected residual value model and the regression coefficients; and

communicating an updated web page with updated final price for the used vehicle to the client device.

3. The system of claim 1, wherein, in the front end online process, the bin determined from the plurality of bins for the used vehicle is defined as a group of vehicles in the historical sales transaction data that are of same year or generation, make, model, or trim as the used vehicle or that have a mileage, condition, or locale that is similar to that of the used vehicle.

4. The system of claim 3, wherein the front end online process further comprises:

determining an average days-to-sell based on historical transactions in the historical sales transaction data corresponding to the group of vehicles in the bin determined from the plurality of bins for the used vehicle; and

including the average days-to-sell on a price report for the used vehicle.

5. The system of claim 1, wherein the front end online process further comprises:

determining whether there is a sufficient number of historical transactions within a predetermined distance of the locale for the given-year, make, and model in the user input data received over the network or Internet through the user interface from the client device; and

if the number of historical transactions is sufficient, determining an average price from the historical transactions within the predetermined distance of the locale for the given year, make and model, wherein generating the responsive web page comprises generating a web page configured to display a curve with the average price and the final price, the final price movable in the web page relative to the average price responsive to the controls;

if the number of historical transactions is insufficient, performing a clustering process to include additional historical transactions involving vehicle models most similar to the used vehicle.

6. The system of claim 1, wherein the front end online process further comprises:

normalizing the final price for the used vehicle based on a price ratio of vehicles within the predetermined distance of the locale.

7. A computer program product comprising a non-transitory computer-readable medium storing instructions executable by a processor for running a vehicle data application on a server computer for:

receiving or collecting, through application interfaces of the vehicle data application, data from distributed data sources communicatively connected to the server computer via a network or Internet, the data from the distributed data sources including historical sales transaction data comprising;

individual historical sales transaction data for a plurality of sold vehicles having a plurality of vehicle configurations sold in a plurality of regions,

vehicle residual value data for the plurality of vehicle configurations, and

geo-specific data for the plurality of vehicle configurations,

wherein each vehicle configuration in the plurality of vehicle configurations comprises a year, make and model, and

wherein the individual historical sales transaction data for the plurality of sold vehicles comprises sale prices and vehicle specific usage or maintenance data for the plurality of sold vehicles;

storing, in a data store, the historical sales transaction data, the vehicle residual value data, and the geo-specific data for the plurality of vehicle configurations;

performing a process divided into a back end process and a front end online process, the back end process performed at a time interval and asynchronously to the front end online process,

wherein the back end process performed by the vehicle data application on the server computer comprises;

generating exponential decay estimate models for the plurality of vehicle configurations, each exponential decay estimate model modelling predicted residual values for a corresponding vehicle configuration for a given age input based on fitting an exponential decay curve to residual value data associated with the corresponding vehicle configuration;

generating residual value models for the plurality of vehicle configurations, each residual value model modelling residual value for a corresponding vehicle configuration at a transaction level as a function of predicted residual values for sold vehicles of the corresponding vehicle configuration determined by applying the exponential decay estimate model, thus generated in the back end process by the vehicle data application for the corresponding vehicle configuration, to individual historical sales transaction data for the corresponding vehicle configuration, regions associated with the individual historical sales transaction data for the corresponding vehicle configuration, and vehicle specific usage or maintenance data from the individual historical sales transaction data for sold vehicles of the corresponding configuration;

applying clustering rules to classify the historical sales transaction data for the plurality of sold vehicles into a plurality of bins based on vehicle characteristics and regions;

after classifying the historical sales transaction data for the plurality of sold vehicles into the plurality of bins, constructing research datasets for vehicle pricing, the research datasets comprising temporally-weighted historical data, geo-specific socioeconomic data, vehicle-specific attributes, vehicle usage or maintenance data, and days-to-sell vehicle data;

constructing price ratio models based on a multivariable regression analysis, the price ratio models modelling a predicted price ratio for a vehicle i in a bin q of the plurality of bins as a function of a mean average price ratio for the bin q, a set of regression variables representing the vehicle-specific attributes, the vehicle usage or maintenance data, the geo-specific socioeconomic data, and the days-to-sell vehicle data, wherein a price ratio model is fitted to data in a dataset of the research datasets thus constructed in the back end process by the vehicle data application to find regression coefficients of the price ratio model;

storing, in the data store, the exponential decay estimate models, the residual value models, the research datasets, and the price ratio models generated in the back end process performed by the vehicle data application on the server computer; and

wherein the front end online process performed by the vehicle data application on the server computer comprises;

presenting a user interface on a client device over the Internet, the user interface comprising a web-based form with fields for entry of user input data through a client browser running on the client device;

receiving the user input data via the web-based form from the client device over the network or Internet through a network communication channel, the user input data comprising information about a used vehicle in a locale, including year, make, model, trim, mileage, and condition of the used vehicle;

receiving or obtaining socio-economic data specific to the locale from the user input data, the socio-economic data specific to the locale comprising income data, census data, home prices, or workforce participation data specific to the locale;

applying a set of rules to the locale and the year, make, model, trim, mileage, and condition of the used vehicle to determine a bin from the plurality of bins for the used vehicle;

constructing regression variables for the used vehicle based on the year, make, model, trim, mileage, and condition of the used vehicle and the socio-economic data specific to the locale;

determining an average price ratio of the determined bin;

applying the price ratio model using the average price ratio for the determined bin, the constructed regression variables for the used vehicle, and the regression coefficients of the price ratio model to determine a price ratio for the used vehicle;

selecting a residual value model for the used vehicle from the plurality of residual value models generated in the back end process by the vehicle data application based on the year, make, model, trim, mileage, and condition of the used vehicle;

determining a depreciated value of the used vehicle using the selected residual value model;

determining a final price for the used vehicle as a function of the depreciated value of the used vehicle and the price ratio for the used vehicle;

generating a responsive web page including the final price for the used vehicle and user interface controls linked to at least one of the regression variables for the used vehicle, the user interface controls configured to allow real-time adjustment of the linked regression variable in the responsive web page; and

communicating the responsive web page to the client device, wherein the responsive web page is generated and communicated to the client device in response to the vehicle data application receiving the user input data from the client device.

8. The computer program product of claim 7, wherein the front end online process further comprises:

receiving, through the user interface on the client device over the network or Internet, updated user input data comprising information about the used vehicle in the locale, including year, make, model, trim, mileage or condition;

adjusting the regression variables constructed for the used vehicle;

adjusting the final price determined for the used vehicle based on the updated user input data using the selected residual value model and the regression coefficients; and

communicating an updated web page with updated final price for the used vehicle to the client device.

9. The computer program product of claim 7, wherein, in the front end online process, the bin determined from the plurality of bins for the used vehicle is defined as a group of vehicles in the historical sales transaction data that are of same year or generation, make, model, or trim as the used vehicle or that have a mileage, condition, or locale that is similar to that of the used vehicle.

10. The computer program product of claim 9, wherein the front end online process further comprises:

determining an average days-to-sell based on historical transactions in the historical sales transaction data corresponding to the group of vehicles in the bin determined from the plurality of bins for the used vehicle; and

including the average days-to-sell on a price report for the used vehicle.

11. The computer program product of claim 7, wherein the front end online process further comprises:

determining whether there is a sufficient number of historical transactions within a predetermined distance of the locale for the given year, make, and model in the user input data received over the network or Internet through the user interface from the client device; and

if the number of historical transactions is sufficient, determining an average price from the historical transactions within the predetermined distance of the locale for the given year, make and model, wherein generating the responsive web page comprises generating a web page configured to display a curve with the average price and the final price, the final price movable in the web page relative to the average price responsive to the controls;

if the number of historical transactions is insufficient, performing a clustering process to include additional historical transactions involving vehicle models most similar to the used vehicle.

12. The computer program product of claim 7, wherein the front end online process further comprises:

normalizing the final price for the used vehicle based on a price ratio of vehicles within the predetermined distance of the locale.

13. A method for determining used vehicle pricing and presenting and updating used vehicle price information in response to site visitor queries, comprising, by a vehicle data system comprising a server computer that provides a web site or web service accessible via Internet and a vehicle data application on the server computer, the server computer configured to respond to site visitor queries from distributed users for vehicle pricing data and generate a responsive interface, the vehicle data application having a data gathering module, an interface module, and a processing module, the interface module comprising application interfaces for distributed data sources communicatively connected to the vehicle data system, the server computer, when running the vehicle data application, performing the steps of:

receiving or collecting, through the application interfaces of the vehicle data application, data from the distributed data sources, the data from the distributed data sources including historical sales transaction data comprising;

individual historical sales transaction data for a plurality of sold vehicles having a plurality of vehicle configurations sold in a plurality of regions,

vehicle residual value data for the plurality of vehicle configurations, and

geo-specific data for the plurality of vehicle configurations,

wherein each vehicle configuration in the plurality of vehicle configurations comprises a year, make and model, and

wherein the individual historical sales transaction data for the plurality of sold vehicles comprises sale prices and vehicle specific usage or maintenance data for the plurality of sold vehicles;

storing, in a data store, the historical sales transaction data, the vehicle residual value data, and the geo-specific data for the plurality of vehicle configurations;

performing a process divided into a back end process and a front end online process, the back end process performed at a time interval and asynchronously to the front end online process,

wherein the back end process performed by the vehicle data application on the server computer comprises;

generating exponential decay estimate models for the plurality of vehicle configurations, each exponential decay estimate model modelling predicted residual values for a corresponding vehicle configuration for a given age input based on fitting an exponential decay curve to residual value data associated with the corresponding vehicle configuration;

generating residual value models for the plurality of vehicle configurations, each residual value model modelling residual value for a corresponding vehicle configuration at a transaction level as a function of predicted residual values for sold vehicles of the corresponding vehicle configuration determined by applying the exponential decay estimate model, thus generated in the back end process by the vehicle data application for the corresponding vehicle configuration, to individual historical sales transaction data for the corresponding vehicle configuration, regions associated with the individual historical sales transaction data for the corresponding vehicle configuration, and vehicle specific usage or maintenance data from the individual historical sales transaction data for sold vehicles of the corresponding configuration;

applying clustering rules to classify the historical sales transaction data for the plurality of sold vehicles into a plurality of bins based on vehicle characteristics and regions;

after classifying the historical sales transaction data for the plurality of sold vehicles into the plurality of bins, constructing research datasets for vehicle pricing, the research datasets comprising temporally-weighted historical data, geo-specific socioeconomic data, vehicle-specific attributes, vehicle usage or maintenance data, and days-to-sell vehicle data;

constructing price ratio models based on a multivariable regression analysis, the price ratio models modelling a predicted price ratio for a vehicle i in a bin q of the plurality of bins as a function of a mean average price ratio for the bin q, a set of regression variables representing the vehicle-specific attributes, the vehicle usage or maintenance data, the geo-specific socioeconomic data, and the days-to-sell vehicle data, wherein a price ratio model is fitted to data in a dataset of the research datasets thus constructed in the back end process by the vehicle data application to find regression coefficients of the price ratio model;

storing, in the data store, the exponential decay estimate models, the residual value models, the research datasets, and the price ratio models generated in the back end process performed by the vehicle data application on the server computer; and

wherein the front end online process performed by the vehicle data application on the server computer comprises;

presenting a user interface on a client device over the Internet, the user interface comprising a web-based form with fields for entry of user input data through a client browser running on the client device;

receiving the user input data via the web-based form from the client device over the network or Internet through a network communication channel, the user input data comprising information about a used vehicle in a locale, including year, make, model, trim, mileage, and condition of the used vehicle;

receiving or obtaining socio-economic data specific to the locale from the user input data, the socio-economic data specific to the locale comprising income data, census data, home prices, or workforce participation data specific to the locale;

applying a set of rules to the locale and the year, make, model, trim, mileage, and condition of the used vehicle to determine a bin from the plurality of bins for the used vehicle;

constructing regression variables for the used vehicle based on the year, make, model, trim, mileage, and condition of the used vehicle and the socio-economic data specific to the locale;

determining an average price ratio of the determined bin;

applying the price ratio model using the average price ratio for the determined bin, the constructed regression variables for the used vehicle, and the regression coefficients of the price ratio model to determine a price ratio for the used vehicle;

selecting a residual value model for the used vehicle from the plurality of residual value models generated in the back end process by the vehicle data application based on the year, make, model, trim, mileage, and condition of the used vehicle;

determining a depreciated value of the used vehicle using the selected residual value model;

determining a final price for the used vehicle as a function of the depreciated value of the used vehicle and the price ratio for the used vehicle;

generating a responsive web page including the final price for the used vehicle and user interface controls linked to at least one of the regression variables for the used vehicle, the user interface controls configured to allow real-time adjustment of the linked regression variable in the responsive web page; and

communicating the responsive web page to the client device, wherein the responsive web page is generated and communicated to the client device in response to the vehicle data application receiving the user input data from the client device.

14. The method of claim 13, wherein the front end online process further comprises:

receiving, through the user interface on the client device over the network or Internet, updated user input data comprising information about the used vehicle in the locale, including year, make, model, trim, mileage or condition;

adjusting the regression variables constructed for the used vehicle;

adjusting the final price determined for the used vehicle based on the updated user input data using the selected residual value model and the regression coefficients; and

communicating an updated web page with updated final price for the used vehicle to the client device.

15. The method of claim 13, wherein, in the front end online process, the bin determined from the plurality of bins for the used vehicle is defined as a group of vehicles in the historical sales transaction data that are of same year or generation, make, model, or trim as the used vehicle or that have a mileage, condition, or locale that is similar to that of the used vehicle.

16. The method of claim 15, wherein the front end online process further comprises:

determining an average days-to-sell based on historical transactions in the historical sales transaction data corresponding to the group of vehicles in the bin determined from the plurality of bins for the used vehicle; and

including the average days-to-sell on a price report for the used vehicle.

17. The method of claim 13, wherein the front end online process further comprises:

determining whether there is a sufficient number of historical transactions within a predetermined distance of the locale for the given year, make, and model in the user input data received over the network or Internet through the user interface from the client device; and

if the number of historical transactions is sufficient, determining an average price from the historical transactions within the predetermined distance of the locale for the given year, make and model, wherein generating the responsive web page comprises generating a web page configured to display a curve with the average price and the final price, the final price movable in the web page relative to the average price responsive to the controls;

if the number of historical transactions is insufficient, performing a clustering process to include additional historical transactions involving vehicle models most similar to the used vehicle.

18. The method of claim 13, wherein the front end online process further comprises:

normalizing the final price for the used vehicle based on a price ratio of vehicles within the predetermined distance of the locale.