Home/Solutions/Catalog matching
Catalog matching

Match every part. Trust every catalog.

Claro decides whether two records are the same product, the same part, or a valid replacement — the call your team makes by hand today.

See it on your data

Why it matters

The attribute that decides a match is rarely in the title

The deciding attribute is almost never in the product title. A2 versus A4, 2RS versus ZZ, a three-millimeter tolerance that changes whether a part fits — these are the details that separate a correct match from a costly one, and they live buried in a spec sheet, not a headline. Teams cover the gap today with side-by-side comparison, one product at a time, which is accurate but does not scale past a few hundred SKUs a week. Claro reads the same attributes a specialist would check, states the ones that drove its decision, and leaves the final call reviewable rather than opaque.

Attribute-level matchingnot just description similarity
A stated reasonfor every match, not a black-box score
Built for crosswalksyour range against competitors, at scale
The problem

Why supplier and internal catalogs never line up

01

A2 or A4, 2RS or ZZ — the attribute that decides a match is rarely in the title.

02

Competitor crosswalks take days of side-by-side spec comparison.

03

A wrong match costs a return; a missed match costs a sale.

How Claro does it

How catalog matching works with Claro

Compare

Every attribute, not just the description.

Score

A confidence score on every match, not a guess.

Explain

The reason for the call, in plain language.

Write back

Matched pairs land in your PIM with their evidence.

Who it is for

Who needs catalog matching

Multi-supplier distributorsRanges where the same part arrives from four suppliers under four different codes.
Marketplaces and aggregatorsPlatforms reconciling seller catalogs against a canonical product list.
Manufacturers building crosswalksTeams mapping their own range against competitor part numbers at scale.
Inputs and outputs

What goes in, what comes back

Reads
Supplier catalogPrice listERP exportBill of materialsCompetitor listingExploded drawing
Returns
Same product / same part / valid replacementConfidence score per pairThe attributes that decided itCrosswalk tableReview queue
Works with
SAPMicrosoft DynamicsAkeneoPimcoreinRiverCustom ERPCSV and API

Claro writes back through files and APIs rather than certified connectors, so this list is a guide, not a limit.

FAQ

Catalog matching: common questions

Do we need GTINs or shared identifiers to match?

No. Most industrial and own-label ranges have no shared identifier, which is exactly the case Claro is built for. Matching runs on attributes, manufacturer and part-number structure, and the supporting documents.

How accurate is the matching?

It depends on the category and how much evidence the source data carries — we do not quote a single headline number that applies to every catalog. Every match comes with a calibrated confidence score, and you set the threshold at which a match is accepted, reviewed or rejected.

What is the difference between "same product", "same part" and "valid replacement"?

Same product means one identical item under two names. Same part means functionally identical for fitment, which may differ in branding or packaging. A valid replacement supersedes the original. Claro returns which of the three applies, because the commercial consequence of each is different.

Does a match write to our ERP automatically?

Only if you configure it to. A link between two catalogs is high blast radius, so by default it lands in the review queue whatever the score.

How does it handle different languages and units?

Units are normalised before comparison — inches to millimetres, litres to millilitres — and descriptions are read in the source language rather than translated first, so meaning is not lost in a round trip.

Build once. Deploy across the catalog. Improve over time.

See it work on your own catalog.

Bring one supplier file and we'll run catalog matching on your real data — matched, classified and reviewable.