Commerce CensusContactRequest access

Home / Industries / Auto parts

Commerce intelligence for auto parts

Auto parts is a fitment problem before it is a price problem. A part number that fits three model years and not a fourth is a different product, and matching that ignores fitment produces confidently wrong comparisons.

MPN + fitmentIdentityOEM vs aftermarketSplitFalse-positive matchesRisk

Auto parts

Fitment as identity

Manufacturer part numbers carry fitment metadata as attributes, so a comparison never crosses a fitment boundary just because two titles look similar.

Auto parts

OEM and aftermarket stay apart

These are different markets at different price points. They are matched to separate canonical products rather than pooled into a single misleading spread.

Endpoints that do this

Where it lands in the API.

POST /v1/match

Bulk match

Send up to 10,000 of your own rows and get canonical ids back, each with a method and a confidence — including the ones we refuse to guess at.

→ reference
POST /v1/resolve

Resolve

GTIN, UPC, EAN, ASIN, seller SKU, product link or plain language into one canonical product and every listing that carries it.

→ reference
GET /v1/products/{id}/offers

Offers

Every live offer for a product, with seller identity, landed cost, condition and the provenance of each value.

→ reference

Questions

Common questions.

How are ambiguous part matches handled?

Flagged rather than guessed. Bulk match returns review: true on anything below the trust threshold, because a false positive here sells someone the wrong part.