How to match a supplier catalog to Amazon ASINs
Match a supplier catalog to Amazon ASINs by identifier precedence: GTIN family first, MPN plus brand second, brand and title last. 10-15% no-match is normal.
Last reviewed 6 min readSupplyForge LLC
Match a supplier catalog to Amazon ASINs by identifier precedence: GTIN-family barcodes first, manufacturer part number plus brand second, and brand plus title only as a last resort.
A wholesale price list arrives with anywhere from a few hundred to fifty thousand rows, and every number downstream of the match — ROI, restriction status, restock quantity, purchase order line cost — inherits whatever error the match step introduced. Matching is not a single lookup. It is a ranked sequence of attempts, each carrying different certainty, and the discipline is stopping at the first method that returns a confident answer rather than falling through to a weaker one that returns something.
The identifier precedence ladder
The identifier precedence ladder ranks match methods by how much inference each one requires, running from a supplier-supplied ASIN, which requires none, down to brand and title, which is entirely inference.
| Match method | What the supplier row must carry | Why it ranks here | Main failure mode |
|---|---|---|---|
| Direct ASIN or Amazon URL | An ASIN or a full Amazon product link | The supplier has named the listing; nothing is inferred | Supplier pasted a variation parent or a competitor's listing |
| GTIN family: UPC, EAN, ISBN, GTIN-14 | A barcode whose check digit validates | Amazon indexes listings against these identifiers directly | Leading zero stripped in a spreadsheet; case-pack barcode sent instead of the consumer unit |
| MPN or model number plus brand | A part number and a brand string | Part numbers are unique inside a brand, never across brands | Two brands use the same part number; brand spelled differently on Amazon |
| Brand plus title | A brand and a description | Nothing structured is left to match on | Silently matches a different size, color, or pack count |
Each step down the ladder buys coverage at the cost of certainty. Forge Command matches on UPC, EAN, ISBN, GTIN, MPN or model number, brand plus title, and direct Amazon URLs, and a single wholesale row can carry links to more than one ASIN — which matters when one supplier SKU legitimately sells under both a standalone listing and a multipack.
Why brand and title matching is a last resort
Brand and title matching is a last resort because an Amazon title is a marketing string rather than an identifier, and two products that differ only in size, color, or pack count usually share most of their words.
The dangerous case is not the obvious miss. It is the near miss that looks correct: a single-unit wholesale cost matched to a listing that sells a six-pack. The cost is per unit, the sale price is per pack, and the resulting ROI looks extraordinary. That number then propagates into a shortlist, a purchase order, and a buying decision, and nothing in the arithmetic ever flags it. The margin trap in multipack listings is the most expensive single consequence of a loose title match. Treat a title-only match as a candidate that a person must open the listing to confirm, never as an input to an automated reorder.
Reviewing the ambiguous middle
The ambiguous middle is the set of rows that produced a candidate ASIN without producing certainty, and it is the only part of a catalog match that justifies human time.
Split the output into three buckets rather than two. Rows matched on a check-digit-valid GTIN or a supplier-supplied ASIN are accepted without review. Rows matched only on brand plus title, or on a bare part number with no brand agreement, go to review. Rows with no candidate at all go to the no-match pile and are diagnosed separately, because the reasons a supplier row fails to match are mostly not fixable by trying harder.
Then review in descending order of exposure, not row order. Sort the review bucket by the money you would actually commit — units you would buy multiplied by unit cost — and work down. A wrong match on a row you will never purchase costs nothing. A wrong match on the row with the highest apparent ROI is the one that gets bought first.
Reading your own match rate as a diagnostic
A match rate is a diagnostic of how aggressive your matcher is, not only a measure of catalog quality, and both a low number and a high number carry information.
On a real wholesale catalog, roughly 10 to 15 percent of rows end with no Amazon match, and Forge Command treats that band as normal rather than as a failure. A rate far below that band is rarely good news. It usually means the matcher fell through to brand-and-title matching on rows that had no usable identifier, converting unmatched rows into wrong matches, which is strictly worse than leaving them unmatched. A suspiciously round imported row count — exactly 20,000, exactly 10,000 — almost always means an import limit truncated the file rather than the supplier happening to sell that many items. Check the row count against the source file before you interpret any match percentage at all.
How Forge Command matches a supplier catalog
Forge Command imports a supplier catalog of tens of thousands of rows, matches each line to an Amazon ASIN, and enriches the matches through Amazon SP-API in batches of 20 identifiers per request across the catalog, pricing, and fee endpoints.
A cloud worker drains every tenant's pending queue using that tenant's own Amazon credentials, so quota is not pooled between customers. Each matched row gets a derived sellability status computed from raw SP-API restriction fields rather than stored — Approved to Sell, Applied to Sell, Restricted, or Not Checked — so the status cannot drift out of date. For a one-off file, the bulk scan accepts a pasted or uploaded CSV or Excel price list without a connector build; for a recurring feed there are three no-code supplier integration tiers, covering PDF price lists, API integrations, and website logins.
The honest cost of this: Forge Command requires your own Amazon SP-API credentials before the catalog engine does anything at all. There is real setup work between signing up and matching your first file.
Match quality has to travel with the margin number
Match confidence has to travel with a row all the way to the buy decision, because a margin computed against the wrong ASIN is not a slightly wrong number, it is a fictional one.
Record which method matched each row and keep that alongside the ROI, so that a title-derived match can be excluded from automated actions later. Re-check the match whenever a supplier changes a SKU number or reissues a catalog, because suppliers reuse SKU codes for successor products more often than they announce it. Forge Command computes margin from one shared function, so the ROI shown on the Wholesale tab, on the Products tab, and on a purchase order line can never disagree with each other — which is the right property to have, and also means that when the underlying ASIN is wrong, all three agree on the same wrong answer. Consistency is not verification. Once the match is trusted, the next step is scanning the price list for the profitable rows.
Frequently asked questions
Can I match a supplier catalog with no UPC column
Yes. Matching can run on MPN plus brand, brand plus title, or direct Amazon URLs when no barcode is supplied. Expect a lower confident-match rate and a much larger review bucket, because every one of those methods requires inference that a valid barcode does not.
Should one wholesale row ever link to more than one ASIN
Yes. A single supplier SKU often sells under several Amazon listings — a standalone unit, a multipack, and sometimes a renewed or refurbished listing. Forge Command allows a single wholesale row to carry links to multiple ASINs rather than forcing a one-to-one relationship.
What no-match rate should I expect from a distributor file
Around 10 to 15 percent of rows on a real wholesale catalog will not match an Amazon ASIN, driven by obscure SKUs, case-pack items, and products genuinely not listed on Amazon. Rates well outside that band usually indicate a mapping or import problem rather than a catalog problem.
Does Forge Command use my own Amazon API credentials
Yes. Every tenant supplies its own API keys, stored encrypted at rest with Fernet, so your Amazon SP-API and Keepa quota belongs to your account rather than to a shared pool. That is also the setup cost: no credentials, no catalog matching.