Wholesale/trade-in pricing system, method and computer program product therefor

• US 10,504,159 B2
• Filed: 09/19/2013
• Issued: 12/10/2019
• Est. Priority Date: 01/29/2013
• Status: Active Grant

• Current Assignee: TrueCar, Inc.
• Original Assignee: TrueCar, Inc.
• Inventors: Laughlin, Isaac Lemon, Liu, Xingchu, Semeniuk, Mikhail, Swinson, Michael D.
• Primary Examiner(s): Abrahamson, Amanda C
• Assistant Examiner(s): Lindsey, III, Johnathan J
• Application Number:
  US14/031,930
• Publication Number:
  US 20140214696A1
• Time in Patent Office: 2,273 Days
• Field of Search: 705306, 705 735, 705400
• CPC Class Codes: G06Q 30/0278 Product appraisal
• CPC Class Codes: G06Q 30/0283 Price estimation or determination

Abstract

Systems, methods and computer program products that determine wholesale prices for durable goods such as used vehicles. In one embodiment, a system includes a computer processor, a network interface, a data storage device, and an output device. The system collects and stores, via the network interface, historical sales information from external data sources. A retail price for a target vehicle is determined based on the collected information. Condition information for the vehicle is collected and used to adjust the retail price to arrive at a wholesale price, which is output to a user. The system may also adjust the price based on repair costs associated with identified vehicle conditions. The system may collect cost information associated with a wholesale sale transaction and deduct these costs and a profit margin from the wholesale price to produce a trade-in price.

Claims

1. A system for determining pricing information, the system comprising:

a computer processor;

a network interface coupled to the processor;

a data storage device coupled to the processor, the data storage device storing a retail pricing model;

an output device coupled to the processor;

wherein the processor is configured to;

collect, via the network interface, historical wholesale transaction records for a set of wholesale transaction for a set of vehicles from data sources external to the system and store the collected historical wholesale transactions on the data storage device, wherein each wholesale transaction includes a sale date, a sale price, a sale region, a vehicle identification and a set of vehicle attributes for a corresponding vehicle, the set of vehicle attributes for a corresponding vehicle comprising a vehicle year, make and model;

provide, over a network, a set of user interface pages with controls to collect vehicle condition information for the vehicles in the set of vehicles, the vehicle condition information comprising first indicators for a plurality of conditions, the plurality of conditions comprising a title condition, a paint condition, a body condition and a windshield condition, and including free-form text;

in a backend process;

apply a first set of rules to bin vehicles in the set of vehicles into a plurality of bins based on make, model, year and region;

extracting first indicators for the plurality of conditions from the free-form text by parsing the free-form text to determine one or more textual condition terms in the free-form text and transforming the one or more textual condition terms determined from the free-form text to first indicators for the plurality of conditions;

classifying the first indicators into one of a set of clusters of second indicators representing the plurality of conditions, wherein the set of clusters is constructed utilizing a machine learning model trained to semantically analyze the free-form text using a training set correlating condition terms to the set of clusters;

enhancing the historical transaction records with the second indicators for the plurality of conditions associated with each historical transaction record;

map the second indicators for the plurality of conditions, to values for a set of condition features, wherein mapping the second indicators for the plurality conditions to values for the set of condition features comprises;

assigning each vehicle in the set of vehicles a numerical value for each feature in the set of condition features;

generate a price ratio model that models price ratio as a function of a set of variables that represent at least the set of vehicle attributes and the set of condition features for the vehicles in the vehicle set, the price ratio model comprising a set of regression coefficients applied to the set of variables, the set of regression coefficients based on a fit of the price ratio model to the historical wholesale transaction and set of condition features;

in a frontend process;

provide a web page to a user device, the web page having one or more input fields for a user to provide a user specified vehicle configuration;

receive, based on user interaction with the web page, a request for a wholesale price of a user specified target vehicle of interest from the user device, the request comprising a set of target vehicle attributes, condition information for the target vehicle and a geographical location, wherein the condition information for the target vehicle indicates conditions associated with the target vehicle and the target vehicle attributes include a target vehicle make, model and year;

determine a retail price for the target vehicle based on the retail pricing model;

apply a second set of rules to determine a bin for the target vehicle from the plurality of bins based on the target vehicle make, model and year and geographical location;

determine from a set of historical records containing wholesale transaction data, vehicles in the set of vehicles associated with the bin;

for each vehicle associated with the bin, determine a predicted retail price based on the retail pricing model;

determine an average ratio of wholesale price to predicted retail price for the bin;

determine a set of target vehicle condition features from the set of condition features based on the condition information for the target vehicle, the set of target vehicle condition features representing conditions associated with the target vehicle;

apply the price ratio model with the set of regression coefficients pre-calculated in the backend process to dynamically generate the wholesale price for the target vehicle of interest as a function of the retail price of the target vehicle, the average ratio of wholesale price to predicted retail price for the bin, the set of target vehicle attributes and the set of target vehicle features representing conditions associated with the target vehicle;

respond to the request in real-time by generating a response interface and providing the response interface to the user device, the response interface configured to reconfigure a user interface at the user device to present the dynamically generated wholesale price for the target vehicle of interest.

2. The system of claim 1, wherein the processor is configured to, in the frontend process:

set variables in the price ratio model for the corresponding ones of the target vehicle features to the associated numeric values when applying the price ratio model.

3. The system of claim 1, wherein the processor is further configured to:

collect wholesale sale transaction cost information from one or more data sources;

determine a trade-in value of the target vehicle based on the wholesale price and the wholesale sale transaction cost information; and

output the trade-in value to the user via the output device.

4. The system of claim 3, wherein the processor is configured to:

collect wholesale sale transaction cost information from a plurality of data sources that are associated with different auction locations;

determine wholesale sale transaction costs associated with the different auction locations;

determine separate trade-in values for each of the different auction locations; and

output the trade-in values for each of the different auction locations to the user via the output device.

5. The system of claim 1, wherein the processor is configured to:

retrieve the wholesale transactions from one or more wholesale auction databases.

6. The system of claim 1, wherein the processor is configured to provide the set of user interface pages to a user input device and wherein condition information input via the set of user interface pages is received by the processor and stored in the data storage device.

7. A method for determining pricing information, the method comprising:

collecting, by a server computer having a processor, a memory and storing a retail pricing model, historical wholesale transaction records for a set of wholesale transaction for a set of vehicles from data sources communicatively connected to the server computer, wherein each wholesale transaction includes a sale date, a sale price, a sale region, a vehicle identification a set of vehicle attributes for a corresponding vehicle, the set of vehicle attributes for a corresponding vehicle comprising a vehicle year, make and model;

storing, by the server computer, the collected historical wholesale transactions on a data storage device;

providing, by the server computer, a set of user interface pages with controls to collect vehicle condition information for the vehicles in the set of vehicles, the vehicle condition information comprising first indicators for a plurality of conditions, the plurality of conditions comprising a title condition, a paint condition, a body condition and a windshield condition;

implementing a backend process by the server computer, the backend process comprising;

applying a first set of rules to bin vehicles in the set of vehicles into a plurality of bins based on make, model, year and region;

extracting first indicators for the plurality of conditions from the free-form text by parsing the free-form text to determine one or more textual condition terms in the free-form text and transforming the one or more textual condition terms determined from the free-form text to first indicators for the plurality of conditions;

classifying the first indicators into one of a set of clusters of second indicators representing the plurality of conditions, wherein the set of clusters is constructed utilizing a machine learning model trained to semantically analyze the free-form text using a training set correlating condition terms to the set of clusters;

enhancing the historical transaction records with the second indicators for the plurality of conditions associated with each historical transaction record;

mapping the second indicators for the plurality of conditions to values for a set of condition features, wherein mapping the indicators for the plurality conditions to values for the set of condition features comprises assigning each vehicle in the set of vehicles a numerical value for each feature in the set of condition features;

generating a price ratio model that models price ratio as a function of a set of variables that represent at least the set of vehicle attributes and the set of condition features for the vehicles in the set of vehicles, the price ratio model comprising a set of regression coefficients applied to the set of variables, the set of regression coefficients based on a fit of the price ratio model to the historical wholesale transactions and set of condition features;

implementing a frontend process by the server computer, the frontend process comprising;

providing a web page to a user device, the web page having one or more input fields for a user to provide a user specified vehicle configuration;

receiving, by the server computer, a request based on user interaction with the web page for a wholesale price of a user-specified target vehicle of interest, the request comprising a set of target vehicle attributes, condition information for the target vehicle and a geographical location, wherein the condition information for the target vehicle indicates conditions associated with the target vehicle and the target vehicle attributes include a target vehicle make, model and year;

determining by the server computer, a retail price for the target vehicle based on the retail pricing model;

applying a second set of rules to determine a bin for the target vehicle from the plurality of bins based on the target vehicle make, model and year and geographical location;

determining from a set of historical records containing wholesale transaction data, vehicles from the set of vehicles associated with the bin;

for each vehicle associated with the bin, determining a predicted retail price based on the retail pricing model;

determining an average ratio of wholesale price to predicted retail price for the bin;

determining a set of target vehicle features from the set of condition features, the set of target vehicle features corresponding to the condition information and representing conditions associated with the target vehicle;

applying the price ratio model using the set of regression coefficients pre-calculated in the backend process to dynamically generate a wholesale price for the target vehicle of interest as a function of the retail price of the target vehicle, the average ratio of wholesale price to retail price for the bin, the set of target vehicle attributes and the set of target vehicle features representing conditions associated with the target vehicle;

responding to the request in real-time by generating a response interface and providing the response interface to the user device, the response interface configured to reconfigure a user interface at the user device to present the dynamically generated wholesale price for the target vehicle of interest.

8. The method according to claim 7, further comprising:

setting variables in the price ratio model for the corresponding ones of the target vehicle features to the associated numeric values when applying the price ratio model.

9. The method according to claim 7, further comprising:

collecting wholesale sale transaction cost information from one or more data sources'"'"'

determining a trade-in value of the target vehicle based on the wholesale price and the wholesale sale transaction cost information; and

outputting the trade-in value to the user via the response interface.

10. The method according to claim 9, further comprising:

collecting wholesale sale transaction cost information from a plurality of data sources that are associated with different auction locations;

determining wholesale sale transaction costs associated with the different auction locations;

determining separate trade-in values for each of the different auction locations; and

outputting the trade-in values for each of the different auction locations to the user via the response interface.

11. The method according to claim 7, further comprising:

retrieving the historical wholesale transactions from one or more wholesale auction databases.

12. The method according to claim 7, further comprising:

providing to a user input device the set of user interface pages, wherein condition information input via the set of user interface pages is stored in the data storage device.

13. A computer program product comprising at least one non-transitory computer-readable storage medium storing computer instructions that are translatable by a server computer to perform:

implementing by the server computer that stores a retail pricing model, a backend process comprising;

collecting historical wholesale transaction records for a set of wholesale transaction for a set of vehicles from data sources communicatively connected to the server computer, wherein each wholesale transaction includes a sale date, a sale price, a sale region, a vehicle identification a set of vehicle attributes for a corresponding vehicle, the set of vehicle attributes for a corresponding vehicle comprising a vehicle year, make and model;

storing the collected historical wholesale transactions on a data storage device;

providing, by the server computer, a set of user interface pages with controls to collect vehicle condition information for the vehicles in the set of vehicles, the vehicle condition information comprising first indicators for a plurality of conditions, the plurality of conditions comprising a title condition, a paint condition, a body condition and a windshield condition;

applying a first set of rules to bin vehicles in the set of vehicles into a plurality of bins based on make, model, year and region;

extracting first indicators for the plurality of conditions from the free-form text by parsing the free-form text to determine one or more textual condition terms in the free-form text and transforming the one or more textual condition terms determined from the free-form text to first indicators for the plurality of conditions;

classifying the first indicators into one of a set of clusters of second indicators representing the plurality of conditions, wherein the set of clusters is constructed utilizing a machine learning model trained to semantically analyze the free-form text using a training set correlating condition terms to the set of clusters;

enhancing the historical transaction records with the second indicators for the plurality of conditions associated with each historical transaction record;

mapping the second indicators for the plurality of conditions, including the indicators determined from the free-form text, to values for a set of condition features, wherein mapping the second indicators for the plurality conditions to values for the set of condition features comprises assigning each vehicle in the set of vehicles a numerical value for each condition feature in the set of condition features;

generating a price ratio model that models price ratio as a function of a set of variables that represent at least the set of vehicle attributes and the set of condition features and the price ratio model comprising a set of regression coefficients applied to the set of variables, the set of regression coefficients based on a fit of the price ratio model to the historical wholesale transactions and set of condition features;

implementing a frontend process by the server computer, the frontend process comprising;

providing a web page to a user device, the web page having one or more input fields for a user to provide a user specified vehicle configuration;

receiving, by the server computer, a request based on user interaction with the web page for a wholesale price of a user-specified target vehicle of interest, the request comprising a set of target vehicle attributes, condition information for the target vehicle and a geographical location, wherein the condition information for the target vehicle indicates conditions associated with the target vehicle and the target vehicle attributes include a target vehicle make, model and year;

applying a second set of rules to determine a bin for the target vehicle from the plurality of bins based on the target vehicle make, model and year and geographical location;

determining from a set of historical records containing wholesale transaction data for vehicles in the set of vehicles, vehicles from the set of vehicles associated with the bin;

for each vehicle associated with the bin, determining a predicted retail price based on the retail pricing model;

determining an average ratio of wholesale price to predicted retail price for the bin;

determining a retail price for target vehicle of interest based on the retail pricing model;

determining a set of target vehicle features from the set of condition features, the set of target vehicle features corresponding to the condition information for the target vehicle and representing conditions associated with the target vehicle;

applying the price ratio model using the set of regression coefficients pre-calculated in the backend process to dynamically generate the wholesale price for the target vehicle as a function of the retail price of the target vehicle, the average ratio of wholesale price to retail price for the bin, the set of target vehicle attributes and the set of target vehicle features representing conditions associated with the target vehicle; and

responding to the request in real-time by generating a response interface and providing the response interface to the user device, the response interface configured to reconfigure a user interface at the user device to present the dynamically generated wholesale price for the target vehicle of interest.

14. The computer program product of claim 13, wherein the computer instructions are further translatable by the server computer to perform:

setting variables in the price ratio model for the corresponding ones of the target vehicle features to the associated numeric values when applying the price ratio model.

15. The computer program product of claim 13, wherein the computer instructions are further translatable by the server computer to perform:

collecting wholesale sale transaction cost information from one or more data sources;

determining a trade-in value of the target vehicle based on the wholesale price and the wholesale sale transaction cost information; and

outputting the trade-in value to the user via the response interface.

16. The computer program product of claim 15, wherein the computer instructions are further translatable by the server computer to perform:

collecting wholesale sale transaction cost information from a plurality of data sources that are associated with different auction locations;

determining wholesale sale transaction costs associated with the different auction locations;

determining separate trade-in values for each of the different auction locations; and

outputting the trade-in values for each of the different auction locations to the user via the response interface.

17. The computer program product of claim 13, wherein the computer instructions are further translatable by the server computer to perform:

providing to a user input device the set of user interface pages, wherein condition information input via the set of user interface pages is stored in the data storage device.