Product Data Cleanup
Product data cleanup corrects the identifier and attribute fields that determine whether products are eligible on Google Shopping, Meta catalogues and marketplaces. The most damaging problems look fine in the admin: barcode fields filled with timestamps or internal IDs from a migration.
Prices shown are starting points. The final cost depends on the size and scope of your project, and we confirm it in writing before any work begins.
Migrated stores are the usual patients. An import fills the barcode field with a timestamp or a row ID, so every product appears to have a barcode. It fails GTIN validation, and because the field is populated, the brand-plus-MPN fallback never engages either: leaving the product with no valid identifier at all.
We validate identifiers properly, correct or clear them, populate brand and MPN where GTIN genuinely does not exist, and standardise attributes so the same concept is not spelled four ways across the catalogue.
What's included
- GTIN and barcode validation with checksum verification
- Junk identifier detection and correction
- Brand and MPN population where GTIN is absent
- identifier_exists flag set correctly
- Attribute standardisation across the catalogue
- SKU convention and duplicate resolution
- Category and product type mapping
- Before-and-after eligibility report
Questions people actually ask
How do I know if my barcodes are junk?
Sort the barcode column and look at it. Real GTINs are 8, 12, 13 or 14 digits with a valid check digit. If you see 13-digit numbers that all start with the same timestamp-like prefix, or values matching your product IDs, they are import artefacts.
Should I just delete invalid barcodes?
For products that genuinely have no GTIN, yes: clear the field and set identifier_exists to false, then make sure brand and MPN are populated. For products that do have a real GTIN, find and enter the correct one; clearing it costs match quality.
Does this affect anything besides Google Shopping?
Yes, Meta catalogues, Amazon and Flipkart listings, and price comparison sites all read the same identifiers. Fixing it at source improves every channel simultaneously.