How to transform product data for marketplaces
Turn ERP, supplier and webshop fields into marketplace-ready attributes without breaking your master catalog. ChannelDock’s PIM Transformations help multichannel sellers reshape titles, variants, units, identifiers and compliance fields before feeds go to Amazon, bol.com, Zalando, OTTO, Kaufland and Google Shopping.
units, values, casing
per marketplace rule
before feeds fail
Product transformation matters because marketplaces no longer accept “roughly correct” catalog data. Google Merchant Center warns that incorrect categories, GTINs, variant attributes and website/feed conflicts can suppress or disapprove products; Amazon has required hundreds of product-type attributes since 2023; Kaufland exposes required, optional and conditional attributes through its API; and OTTO marks legally relevant attributes as mandatory.
Prerequisites: what you need before transforming fields
Transformation rules only work when the source data, channel requirements and ownership model are clear. Otherwise rules become hidden spreadsheet logic that nobody trusts.
Export the exact fields from ERP, Shopify, WooCommerce, supplier files and current marketplace exports. Keep identifiers, attributes, images, descriptions, prices and stock separate.
Collect field names, accepted values, length limits, unit formats, required attributes and conditional rules for each marketplace category you sell in.
Assign one owner for mappings and transformations. Product, marketplace and operations teams can request changes, but one person should approve rules that affect every feed.
Step-by-step: transform product data for marketplace feeds
Use this workflow when marketplace exports keep failing, when supplier files use inconsistent formats, or when the same product needs different outputs for Amazon, bol.com, Zalando, OTTO, Kaufland and Google Shopping.
Audit the feed failures and category rules first
Start with real rejection reports and marketplace documentation, not a generic data-cleaning wishlist. Note whether each issue is a missing value, invalid value, wrong format, length problem, category mismatch, image problem, variant mismatch or compliance field. For Google, common problems include incorrect categories, GTINs and missing variant attributes. For Amazon, the product type controls which attributes matter. For Kaufland and OTTO, legal and category-specific fields can decide whether an item goes live.
Separate source facts from channel output
Create a clean source layer for facts: SKU, EAN/GTIN, brand, product type, dimensions, weight, material, colour, size, safety data, images and parent-child relationships. Then create channel-output fields separately. “Colour” in the source may stay “navy blue”, while Google receives “blue”, bol.com receives a Dutch value, and Zalando receives a controlled option that matches its fashion taxonomy.
Normalize units, casing and forbidden values
Build basic transformations for kilograms to grams, centimetres to millimetres, decimal separators, uppercase/lowercase, empty placeholders, “N/A”, colour synonyms and supplier abbreviations. A small ruleset prevents hundreds of recurring errors. Use PIM Mapping for field alignment and transformations for value cleanup.
Build conditional rules by category and product type
Not every SKU needs the same attributes. Apparel needs size, colour, gender, fit and material. Electronics may need compatibility, power, energy labels and datasheets. Bundles need bundle flags and component logic. Create rules that trigger only for the relevant category, marketplace and product type, then document why the rule exists.
Transform titles without losing variant clarity
Create a title formula per channel: brand + product type + decisive attribute + variant. Keep titles within each marketplace’s length and policy limits. Move secondary details into bullet points, highlights or product details. This matters because mobile results truncate quickly and Amazon, Google and fashion marketplaces increasingly use structured attributes to understand variants.
Preserve parent-child variant relationships
Map parent SKU, variant SKU, item group ID, size, colour, pattern, material and variant option fields consistently. If one marketplace groups by colour and size while another uses model and product version, generate the correct output without changing the underlying family. Validate that all child variants share the same parent identifier where required.
Protect legal and compliance attributes
Mark fields such as GPSR responsible operator, manufacturer, product safety contact, energy certification, age warnings, battery data, material composition and image alt text as controlled data. Transform their format where required, but do not rewrite the substance without approval. Use PIM content quality checks to block feeds when compliance fields are missing.
Preview the output before publishing
Before sending the feed, preview several rows per category and compare source value, transformation rule and final channel value side by side. Include one bestseller, one long-tail SKU, one variant family, one product with missing data and one regulated product. Publish through PIM Feeds only after required fields, field lengths and option values pass validation.
Turn every rejection into a reusable rule
After publishing, review marketplace diagnostics and seller-forum style issues: missing colour, duplicate variants, wrong category, title too long, forbidden image text, invalid GTIN, missing legal attribute. Fix the source or transformation rule, then rerun the feed. The operational goal is fewer manual corrections each cycle, not just today’s upload.
The transformation loop competitors rarely explain
Most PIM and feed-management guides explain mapping. Fewer explain the operating loop: diagnose marketplace feedback, adjust the source or rule, validate outputs, publish, then store the lesson as a reusable transformation. That is what keeps a growing catalog from becoming a spreadsheet repair job.
Good product-data transformation is reversible, documented and channel-specific. If nobody can explain why a value changed, the rule is a future feed error.
Transformation examples by marketplace
Amazon
Create product-type specific outputs. Preserve required attributes, variation themes, bullet points and identifiers. Avoid changing category or item-classification values without checking whether Amazon locks the field after listing creation.
bol.com
Transform Dutch product information, EAN, brand, condition, mandatory category attributes and GPSR fields. If an offer is not publishable, enrich the missing mandatory attributes at the product-data layer.
Zalando / fashion
Focus transformations on colour, size, material, season, images and variant grouping. Fashion feeds need controlled values, not supplier wording copied into every field.
Kaufland / OTTO / Google
Handle legal attributes, category-specific mandatory fields, unit formats, Google product category, item group ID and price/availability consistency. Use conditional rules where attributes only apply to certain categories.
Manual feed edits vs. ChannelDock transformations
The question is not whether someone can repair one export. The question is whether the next supplier import, language launch and marketplace rule change are handled without recreating the same manual work.
Without ChannelDock
- Rules live in spreadsheets, feed apps and marketplace portals with no single owner.
- Teams fix rejected rows manually, so the source record remains wrong.
- Each new channel means another export format, another copy of titles and another drift point.
With ChannelDock
- Mapping, transformations, quality checks and feeds work from one product-data workflow.
- Marketplace-specific outputs are generated without damaging the master catalog.
- Repeated errors become reusable rules for Amazon, bol.com, Zalando, OTTO, Kaufland and Google Shopping.
Common transformation pitfalls
- Changing master data to satisfy one marketplace instead of generating a channel-specific output.
- Mapping field names but forgetting value formats, accepted options, units and language rules.
- Using static defaults for required fields without checking whether they are true for the product.
- Letting manual portal edits drift away from the PIM source record.
- Failing to document why a transformation exists, who approved it and which channels depend on it.
Product data transformation FAQ
Related PIM guides
Use these pages to move from transformation rules to complete marketplace publishing.
Transform marketplace product data without breaking your master catalog
Use ChannelDock PIM to map, transform, validate and publish channel-ready product feeds from one operational source.