Product Catalog Migration Checklist

The short version: Never migrate a dirty catalog. Audit and clean the source data first, map every field to its destination explicitly, run a small test batch through the full pipeline, verify counts and spot-check records, and keep the old catalog intact until the new one has survived a full sales cycle. The migration itself is the easy part; preparation and verification are the job.

1. Before anything moves: audit the source

Export the complete catalog and count it: products, variants, images, categories. These counts are your verification baseline.
Find the duplicates – same product under two SKUs, near-identical titles, zombie listings from old seasons.
List the data quality problems: missing images, empty descriptions, inconsistent attributes (three spellings of the same colour), prices of zero.
Decide what does not migrate. Dead products with no sales and no SEO value are cheaper left behind than cleaned.

2. Clean in the spreadsheet, not in the new platform

Cleanup is fastest and safest in the exported data, where changes are visible, reversible, and reviewable. Dedupe, normalize attributes to one controlled vocabulary, standardize title and description formats, and fix identifiers (SKU, GTIN, MPN) while you can still compare against the live store. Migrating first and cleaning later means doing the same work inside a platform you are still learning, under launch pressure.

3. Map every field explicitly

Write the mapping down: source field → destination field, one row each, including the boring ones (weight, barcode, tax class, SEO title).
Flag fields with no destination. Decide each one: drop it, repurpose a metafield, or restructure. Silence here is how data gets lost.
Handle variants deliberately: option names and values rarely line up one-to-one between platforms.
Plan URLs and redirects alongside the data. Every changed product URL needs a redirect, or the migration quietly burns your search traffic.

4. Test with a batch that can hurt you

Run 50 to 100 products through the entire pipeline first – and choose the awkward ones: multi-variant products, products with many images, grouped or bundled items, the longest descriptions. A test batch of simple products proves nothing. Verify the test batch on the new platform the way a customer would: open the pages, check every variant's price and stock, look at the images, and diff a sample of records field by field against the source.

5. Migrate, then verify like an auditor

Compare totals: products, variants, images in versus out. Counts must reconcile exactly, or the difference must be explained in writing.
Spot-check a random sample (at least 2 to 5 percent) field by field.
Run the feed check: export from the new platform and submit to Google Merchant Center diagnostics – feed errors surface data problems the storefront hides.
Test the money paths: search, filters, add to cart, checkout, and an actual order on the new catalog.

6. After launch: keep the old catalog on ice

Do not delete the source catalog at launch. Keep it intact and accessible for at least one full sales cycle – the missing field or mangled attribute always surfaces in week three, and recovery is trivial if the source still exists. Monitor feed disapprovals, search traffic on migrated URLs, and support tickets mentioning products for the first month; they are your early-warning system.

Planning a migration?

We clean and migrate catalogs between WooCommerce, Shopify, and custom platforms – audited, mapped, and verified, like the 12,000-SKU cleanup in our portfolio.

Catalog Cleanup & Migration Ask a Question

The Product Catalog Migration Checklist: Move Platforms Without Losing Data