MarketCheck Compass Detailed Auto Sales Data

#

Overview

VIN-level records of vehicle sales inferred from North American dealership website listings. Each record describes one vehicle on the day it was sold, with 37 fields covering vehicle identity, mileage, pricing, days on market at sale, seller identity (including mapped stock ticker symbols for publicly traded dealership groups), and geography. Both new and used vehicle sales are included. The event-level structure supports measurement of sales volumes, transaction-price trends, days-to-sale, and sales mix at the vehicle, dealer, dealership-group, brand, or market level.

MarketCheck crawls dealership websites across North America every day and structures the results into VIN-level data on vehicle listings, inventory, sales, and pricing. Its collection covers more than 100,000 dealer websites and observes roughly 6 million unique VINs per day.

Data Description

One row per vehicle on the day it was sold, keyed by pseudo_vin and status_change_date — approximately 334 million rows spanning January 2017 through early July 2026 as of the initial delivery.

Coverage

United States and Canada, 2017 to the present. Coverage spans 100% of franchise new-car dealers and 92% of independent used-car dealers. Four categories of vehicles are not captured: (1) vehicles from OEMs that sell without dealer networks (e.g., Tesla, Rivian, Lucid), which appear only when resold as used vehicles; (2) fleet vehicles, which are not posted on the web until resale; (3) custom factory-direct sales, likewise captured only at resale; and (4) small independent dealers without websites, roughly 8% of independent dealers.

Methodology

Web-scraped daily from franchise and independent dealership websites in the United States and Canada. Vehicle identification numbers are pseudonymized as an MD5 hash (pseudo_vin). A sale is inferred when a vehicle leaves dealer inventory: the sale record appears with a status_change_date approximately seven days after the vehicle's final day in the companion inventory dataset. Provider quality assurance removes implausible rows (odometer readings above 500,000 miles, days on market above 730, prices above $500,000 outside the super-luxury category) and imputes sparse powertrain, drivetrain, and fuel-type nulls from the modal value of matching vehicle cohorts; remaining nulls (under 0.1%, concentrated in discontinued makes) are left in place. Combustion-vehicle fuel-economy outliers are replaced with cohort modes.

Additional Notes

Sales are inferred from listing removal rather than transaction records, and recorded prices reflect the final listed price rather than a verified transaction price. The companion MarketCheck Compass Detailed Auto Inventory Data dataset records the daily listing history of the same vehicles.