Industries / Retail, hospitality and personal services / Retail and e-commerce
Auto parts retail
The auto parts retail industry sells replacement parts and accessories for cars and trucks, from stores, warehouses, delivery vans and websites, to drivers who fit the parts themselves and to the repair shops that fit them for others.
1909
George Pepperdine, a 23-year-old bookkeeper in Kansas City, founded the Western Auto Supply Agency with five dollars and sold parts for the new Ford Model T by mail from his home, with his wife Lena packing the orders. Pepperdine University dates the founding to 1909, and the Kansas City Public Library describes the mail-order start and a September 1909 advertisement.
30
Milestones achieved across 6 domains.
9
To watch for: announced, not yet confirmed outside.
12
Open deliverables. Nobody has shipped them.
In one minute
Where it started
The first cars sold to ordinary families, such as the Model T, needed tires and spare parts that their makers did not stock near most drivers. Western Auto began selling such parts by mail in 1909, and in 1925 independent parts sellers formed NAPA in Detroit to share distribution. For decades a parts counter held racks of paper catalogs, one for each supplier, until electronic catalogs (1984) and shared data standards, TecDoc in Europe (1994) and ACES in North America (2003), recorded which part fits which vehicle.
Where it stands
The Auto Care Association, a US trade group, puts the average age of US cars and light trucks at a record 12.9 years and forecasts US aftermarket sales of $599.7 billion in 2026. Large chains hold stock in tiers: an AutoZone store carries 20,000 to 25,000 items and its biggest hubs up to 100,000, while LKQ's European business holds more than 900,000. Online sellers work alongside the stores, among them Tuhu in China, which fits its orders at 8,008 workshops. Fakes are a growing problem: US Customs and Border Protection seized more than 211,000 counterfeit vehicle parts in fiscal year 2024, almost double the year before.
What is open
Many open jobs are checks that nobody publishes: how often "fits your vehicle" claims are wrong, what parts returned as faulty turn out to be, how long deliveries to repair shops really take, and whether sellers take back old batteries as the law requires. Others are public datasets: a checked table of which part fits which car, a log of counterfeit seizures by part type, and a count of which parts for electric cars can be bought outside the car maker's dealers. AI models that read photos and web pages, browser agents and open vehicle registers put these jobs within reach of a small team.
Domains
AchievedWatch forOpen
Domain
Progress
Last achieved
Watch for
Open
Catalogs and fit
Catalog work records which part fits which vehicle, so that a counter clerk, a mechanic or an online shopper can pick the right one from millions of parts.
5 · 1 · 2
Last achievedData a computer can read2020
Watch forAI-built fitment records
2
openInventory
Inventory work decides which of hundreds of thousands of part numbers to keep in which building, so that a part for almost any car still on the road can reach a store or a repair shop within hours.
5 · 2 · 2
Last achievedCars keep getting older2025
Watch forSixty market hubs by 2027
2
openStores and delivery to shops
This work runs the counters where drivers and mechanics buy parts, and the vans that carry parts from stores and warehouses to repair shops, often within an hour of the order.
4 · 1 · 2
Last achievedShops order on screens2008
Watch forDrones carry parts to shops
2
openOnline sales
Online sales is selling parts to drivers and repair shops by mail order, website, app or marketplace, and getting the right part to the buyer.
4 · 2 · 2
Last achievedOnline orders, workshop fitting2023
Watch forAI shopping guides for tires
2
openReturns and cores
Returns work takes back parts that were the wrong fit, not needed or said to be faulty, and collects the worn parts, called cores, and the old batteries that customers hand in so they can be rebuilt or recycled.
5 · 1 · 2
Last achievedAny seller takes batteries2025
Watch forFaulty parts flown back
2
openFakes and copies
A counterfeit part carries the name of a brand that did not make it, while a copy sold under its own maker's name can be lawful, and this work keeps the first out of shops and online listings and sets the line between the two.
7 · 2 · 2
Last achievedEU frees visible spare parts2024
Watch forQR codes on filters
2
open
Advances to build with
1 research note
Open in auto parts retail
3 of 12 open deliverables
Open · Catalogs and fit
An open, checked fitment set
A free table of 2,000 part-to-vehicle pairs for five everyday parts (oil filters, cabin air filters, wiper blades, front brake pads and batteries) on 100 vehicles common in North America and Europe. For each pair it records what at least three parts makers' own public catalogs say, with the page and date of each reading, and flags where they disagree.
Built with what exists
NHTSA's free vPIC VIN decoder and its downloadable database, for the vehicle list; parts makers' public catalog lookups, read by hand or with a browser agent such as Browser Use where a site's terms allow it (see the Note "Deleting 80% to start over"); Python; CSV files on GitHub.
Open · Inventory
Inspection failures as demand
A public forecast, made from the records of Britain's MOT test (the yearly roadworthiness check), of how many brake, suspension, lighting and tire failures each make, model and age of car will show in the next year of records. Failures are a stand-in for the replacement parts that stores will be asked for. The forecast is scored against that year's records once they are published.
Built with what exists
The Driver and Vehicle Standards Agency's anonymized MOT data, which gives every test since 2005 with make, model, odometer reading and reasons for failure; DuckDB and Python; a gradient-boosting library such as LightGBM, set against a simple "same failure rate as last year" baseline.
Open · Stores and delivery to shops
Clock the shop delivery
A shared log in which independent repair shops record, for every parts order, the time of the order, the arrival time the supplier promised and the time the part arrived, published as delivery times for each kind of supplier and each city.
Built with what exists
A phone form that takes a photo of each delivery slip; a vision-language model to read order numbers and times from the photos, with a person checking each reading; SQLite or a spreadsheet; Python for the statistics.
Where it sits
Broader
