Onboard a New Supplier Range in 24 Hours

A new supplier ships you a spreadsheet of 4,000 SKUs and expects you live by next week. Most catalog teams spend two to six weeks on this: chasing encoding errors, hand-remapping columns, and discovering duplicates only after they hit the storefront. This playbook shows you how to onboard a supplier range fast — mapping, matching, enriching, and publishing a clean, sellable catalog inside a single working day.

Claro sits between your incoming supplier feed and your PIM or ERP as a permanent data layer. It resolves product identity across feeds (so you never create a duplicate you already stock), enriches missing attributes with sourced values, validates identifiers and units in bulk, and writes clean records directly back into your existing systems. Every step below can be run manually, but teams that route supplier feeds through Claro compress this playbook from days to hours because the identity-resolution, enrichment, and write-back steps are automated rather than hand-built.

The outcome: a validated set of canonical product records, each linked to its source file, ready to push into your PIM, ERP, or storefront. Works equally well for an MRO fastener line, a CPG grocery range, a furniture import, or an industrial drives catalog.

Before you start

You need three inputs and one decision. Have the supplier’s raw file (CSV, Excel, or BMEcat), your own target schema, and your matching rules ready. The decision: which fields are mandatory to go live versus nice-to-have. Everything below assumes you batch the work rather than touching records one at a time.

The 24-hour pipeline at a glance

1. 1
   Normalize the incoming file

   Suppliers send broken delimiters, mixed encodings, and merged header rows. Before anything else, clean the file so every column parses. Run it through the CSV fixer to resolve UTF-8 versus Latin-1 issues, stray quotes, and inconsistent separators. Confirm row counts match the supplier’s stated SKU total — a mismatch usually means a delimiter swallowed a column.

2. 2
   Map supplier columns to your schema

   Map every incoming field to a target attribute: their “Art-Nr” to your MPN, their “VPE” to your pack quantity, their free-text “Category” to your taxonomy node. Decide the mandatory set now — typically identifier, title, manufacturer, UOM, and price. See How to Map Supplier Attributes to Your Schema for a repeatable field-mapping approach, and the schema mapping glossary entry if the concept is new to your team.

3. 3
   Validate identifiers and units

   Bad barcodes and ambiguous units stall every launch. Check that GTINs carry valid check digits and that units resolve to a standard (each, metre, litre, kilogram). A furniture range listing “1 box = 2 chairs” needs the pack relationship captured, not flattened. Reject or quarantine rows with invalid identifiers rather than guessing. The unit-of-measure glossary entry explains why unit mismatches are the most common silent data error in supplier feeds.

4. 4
   Match against your existing catalog

   Many “new” supplier items are products you already stock under a different MPN or from another vendor. Run the range against your inventory to find overlaps before you create duplicates. The How to Match Supplier Catalogs to Your Inventory playbook covers blocking and scoring; new genuinely-distinct items pass straight through, while likely matches go to a short review queue. Claro’s identity resolution layer handles this automatically, combining deterministic key matching with fuzzy scoring and routing borderline cases to a human-review queue.

5. 5
   Classify and enrich the gaps

   Assign each product to your taxonomy (ETIM, UNSPSC, or an internal tree) and fill missing attributes — dimensions, material, voltage, compliance flags — from the supplier datasheets. When you pull a spec from a PDF or auto-generate a value, keep the source attached. How to Fill Missing Attributes With Provenance shows how to enrich without inventing data. Data provenance is what lets you defend a value in a customer dispute or roll it back if a supplier corrects it.

6. 6
   Build canonical records and publish

   Merge each validated row into a single golden record per product, with every field traceable to its origin. Push the mandatory-complete records live and hold the rest in a “needs enrichment” bucket so the launch is not blocked by long-tail gaps. Log a per-supplier quality snapshot so you can hold the vendor accountable next time — a supplier data scorecard turns that into a habit.

Before and after: messy vs trusted catalog

Common pitfalls

Other traps that turn a 24-hour onboard into a 24-day one:

• Treating the supplier file as truth. Vendor data is a starting point, not a golden record. A CPG supplier’s “net weight” may be the case weight, not the unit. Validate, do not trust.
• Going live on 100% completeness. Launch the records that meet your mandatory fields and enrich the long tail in parallel. Waiting for every attribute is why ranges sit in a backlog for weeks.
• Losing provenance. If you cannot say where a spec came from, you cannot defend it when a customer disputes it or an AI search engine asks for a source. Keep the link from day one.
• Running a one-off script instead of a repeatable layer. Hand-built CSV scripts break the next time the supplier changes their column order. A canonical data layer preserves the mapping and the matching rules so the second onboard takes minutes, not days.

Related

FAQ