Product data cleanup services — import-ready catalogs.

Duplicate SKUs, near-duplicate variants, and the same product imported twice from two suppliers are the most common reason a catalog won't load cleanly. We identify duplicates by matching across SKU, name, GTIN, and key attributes — then resolve each group with explicit survivorship rules so the most complete, most correct record wins. Nothing is merged blindly: where two records disagree on a value, the conflict is recorded and, if it can't be resolved with confidence, flagged for your call. Every group we collapse is listed in the change-log, so you can see exactly which records merged into which, and why.

Units of measure and barcodes are where small inconsistencies block an import. We normalize every unit of measure to one consistent convention — "33cl", "33 CL" and "330ml" become a single standard value — so quantities, pack sizes, and pricing units are comparable across the whole catalog. Every GTIN is validated against the GS1 check-digit algorithm and normalized to a consistent length (GTIN-8, GTIN-12, GTIN-13, or GTIN-14), with leading zeros handled correctly. A barcode that fails the check digit, or that is shared by two different products, is flagged for your confirmation — never silently corrected, and never dropped.

Inconsistent, free-text, or missing categories make a catalog impossible to browse or filter. You give us the target taxonomy — your own category tree, a marketplace scheme, or the structure your new PIM expects — and we map every product to it. Where source categories are spelled differently, nested differently, or simply blank, we resolve them to the right node in your tree. Products that are genuinely ambiguous, or that have no clear home in your taxonomy, are not forced into a wrong category: they go onto a short decision list for you to confirm. The full before-and-after mapping is recorded in the change-log.

A catalog that loads but is full of blank required fields fails at the next step — the webshop won't publish, the marketplace rejects the feed. We identify the attributes your destination requires, then complete the gaps from the sources you provide: other columns in your export, supplier sheets, spec documents, or your existing master data. We never invent a value to fill a cell. Anything that cannot be completed from a reliable source is surfaced as a short, prioritized decision list — the small set of items that genuinely need someone inside your business to answer — so you spend your time only where a real decision is required.

Mixed formats are silent killers: "1.5" and "1,5", a date that reads as March in one row and the third of the month in the next, prices that mix euros and cents. We normalize dates to ISO 8601 (YYYY-MM-DD), currency amounts to ISO 4217 minor units with the correct currency code, and number formats to one consistent decimal and thousands convention. Trailing whitespace, stray characters, smart quotes, and inconsistent casing are cleaned across every text field. The result is one clean, predictable standard your import or PIM can read without surprises — and every reformatted cell is logged with its original value.

Most PIM migrations and webshop launches stall on dirty product data — not on the software. Pre-PIM data readiness is the work of staging and validating your catalog before it loads, so the migration goes in clean the first time. We consolidate your supplier files and exports into one source, apply all of the cleanup above, and validate the result against the structure and required fields your new PIM or platform expects. You receive an import-ready file mapped to the target schema, plus a findings report that calls out anything the new system will reject — caught here, on a spreadsheet, instead of mid-migration. The other five services above feed straight into this one.