Price scoring for vehicles using pricing model adjusted for geographic region
Vehicle pricing such as used vehicle pricing is improved by supplementing statistical modeling techniques with additional algorithms to accommodate factors such as geography and dealer reputation that do not readily yield to regression analysis or similar tools that might be used to characterize a population.
1. A method for estimating vehicle values in a local geographic region based on national listing data, the method comprising:
obtaining listing data with a first server for vehicles from a number of data sources including one or more data feeds, databases, or websites containing dealer listings coupled to a data network;
creating a pricing model with the first server using the listing data for fair market value of a vehicle type based upon a first data set of vehicles in the listing data obtained for a first geographic region, wherein the pricing model relates vehicle prices to various vehicle attributes for the first data set of vehicles;
receiving a request from a client device over the data network to a web interface of a second server, the request including one or more attributes to narrow or define a search for vehicle listings, and the request further specifying that resulting vehicle listings are local listings in a second geographic region within the first geographic region;
identifying a second data set of vehicles in the listing data from the second geographic region within the first geographic region including vehicles responsive to the request;
calculating a first median for the first data set scored according to the pricing model;
calculating a second median for the second data set scored according to the pricing model;
calculating a median offset between the first median and the second median;
adjusting the pricing model for use when scoring vehicles within the second geographic region by using the median offset to adjust scores calculated for vehicles within the second geographic region using the pricing model for the first geographic region, thereby providing an adjusted pricing model for the second geographic region;
with the second server, scoring a number of vehicles within the second geographic region with the adjusted pricing model, thereby providing scored vehicles;
creating a web page responsive to the request including the scored vehicles within the second geographic region that have been scored according to the adjusted pricing model; and
transmitting the web page including the scored vehicles to the client device over the data network for display at the client device to facilitate comparison shopping for vehicles within the second geographic region by a user of the client device.
2. The method of claim 1 wherein the pricing model is a regression model based upon a plurality of regression parameters including one or more of a model, a year, a mileage, and a trim.
3. The method of claim 1 wherein the pricing model is a regression model based upon a number of regression parameters including one or more of a rental fleet history, a repair history and a flood damage history.
4. The method of claim 1 wherein the first geographic region is a national region.
5. The method of claim 1 wherein the second data set non-exclusively includes vehicle data for a number of metropolitan regions.
6. The method of claim 5 wherein the number of vehicles is a subset of the second data set exclusive to one of the number of metropolitan regions.
7. The method of claim 6 wherein the one of the number of metropolitan regions is a city responsive to a request from the client for listings within the city.
8. The method of claim 1 wherein the second geographic region is a metropolitan region within the first geographic region.
9. The method of claim 1 wherein the pricing model calculates a vehicle price.
10. The method of claim 1 wherein the pricing model calculates a score relative to a standard deviation for the pricing model.
11. A computer program product for estimating vehicle values in a local geographic region based on national listing data, the computer program product comprising computer executable code embodied in a non-transitory computer-readable medium that, when executing on one or more computing devices, performs the steps of:
obtaining listing data with a first server for vehicles from a number of data sources including one or more data feeds, databases, or websites containing dealer listings coupled to a data network;
creating a pricing model with the first server using the listing data for fair market value of a vehicle type based upon a first data set of vehicles in the listing data obtained for a first geographic region, wherein the pricing model relates vehicle prices to various vehicle attributes for the first data set of vehicles;
receiving a request from a client device over the data network to a web interface of a second server, the request including one or more attributes to narrow or define a search for vehicle listings, and the request further specifying that resulting vehicle listings are local listings in a second geographic region within the first geographic region;
identifying a second data set of vehicles in the listing data from the second geographic region within the first geographic region including vehicles responsive to the request;
calculating a first median for the first data set scored according to the pricing model;
calculating a second median for the second data set scored according to the pricing model;
calculating a median offset between the first median and the second median;
adjusting the pricing model for use when scoring vehicles within the second geographic region by using the median offset to adjust scores calculated for vehicles within the second geographic region using the pricing model for the first geographic region, thereby providing an adjusted pricing model for the second geographic region;
with the second server, scoring a number of vehicles within the second geographic region with the adjusted pricing model, thereby providing scored vehicles;
creating a web page responsive to the request including the scored vehicles within the second geographic region that have been scored according to the adjusted pricing model; and
transmitting the web page including the scored vehicles to the client device over the data network for display at the client device to facilitate comparison shopping for vehicles within the second geographic region by a user of the client device.
12. The computer program product of claim 11 wherein the pricing model is a regression model based upon a plurality of regression parameters including one or more of a model, a year, a mileage, a trim, a rental fleet history, a repair history and a flood damage history.
13. The computer program product of claim 11 wherein the first geographic region is a national region.
14. The computer program product of claim 11 wherein the second data set non-exclusively includes vehicle data for a number of metropolitan regions.
15. The computer program product of claim 14 wherein the number of vehicles is a subset of the second data set exclusive to one of the number of metropolitan regions.
16. The computer program product of claim 11 wherein the second geographic region is a metropolitan region within the first geographic region.
17. The computer program product of claim 11 wherein the pricing model calculates a score relative to a standard deviation for the pricing model.