# Migrate Products to Shopify: Mapping SKUs, Variants and Inventory

> Plan a Shopify product migration with field mapping, reliable SKUs, variant tests and inventory reconciliation across stores and multiple locations.

- Author: [Francesco Guiducci](https://ifgdigital.com/en/francesco-guiducci/)
- Published: 2026-11-15 · Updated: 2026-11-15
- Topic: [Migrations and platform choice](https://ifgdigital.com/en/ifgwebdesign/ecommerce/blog/topic/migrazioni/)
- Language: English · Italian version: https://ifgdigital.com/ifgwebdesign/ecommerce/blog/migrazione-prodotti-varianti-shopify/
- Reading time: 6 min
- URL: https://ifgdigital.com/en/ifgwebdesign/ecommerce/blog/migrate-products-to-shopify-skus-variants-inventory/

**Direct answer:** migrating products to Shopify is not simply uploading a spreadsheet. It means translating source records into products, purchasable variants, inventory locations and structured attributes, then proving that what customers can buy matches the original catalogue. A clean import log does not establish that prices, images, options or available quantities are correct.

## Identify the system that owns each field

Some merchants maintain their catalogue inside their original ecommerce platform. Others receive product details from a product information management system (PIM), pricing from an ERP and stock figures from a warehouse management system. These distinctions matter when planning the target.

Create a field inventory covering a stable source ID, SKU, barcode where applicable, product title, description, option names and values, selling price, compare-at price, tax-relevant attributes, image references, category, availability and any custom data. For each field, record the source of truth, proposed Shopify destination, transformation rule and acceptance test.

If the ERP remains responsible for inventory, do not turn a one-time import file into an accidental second inventory master. Define which system may update which values after launch.

## Separate products, purchasable variants and descriptive attributes

A shirt sold in several sizes and colours needs actual purchasable combinations. A fabric composition label, by contrast, might be a structured attribute rather than another variant. A source platform may store these details as product attributes, custom options or extension data; there is no reason to assume they all convert one-to-one.

For every product family, identify the option values that affect what can be added to the cart. Keep specifications that customers need to filter or read in appropriate structured fields, without multiplying sellable combinations unnecessarily.

Shopify's [official product CSV documentation](https://help.shopify.com/en/manual/products/import-export/using-csv) explains how option columns relate to variant rows. The rules have important dependencies: changing variant-related fields while omitting necessary option columns can produce unexpected changes. Test the exact import format on a small store or safe sample before an update to existing products.

## Use SKUs and identifiers for reconciliation

A source product ID and a Shopify product ID are not interchangeable. Keep a mapping between them rather than treating an automatically generated destination ID as if it were the original system's identity.

The SKU is often the operational identifier for a sellable variant, particularly when an ERP or warehouse references it. Identify missing, repeated or non-standard SKU values before the bulk import. A deliberate negative test with a duplicated SKU is more informative than finding duplicates after customers start ordering.

A GTIN, EAN or UPC is a product identification code, not a substitute for every merchant SKU. Transfer genuine codes to the correct destination fields when available; never generate fictitious barcodes just to satisfy an empty column.

Keep a reconciliation table with source record, source variant, destination product, destination variant, SKU and validation status. This also supports future integration debugging.

## Model custom information without burying it in descriptions

Materials, measurements, technical specifications, compatibility notes or production lead times do not always belong in the product description. Depending on use, these can be represented by native Shopify fields, product metafields, variant metafields or reusable metaobjects.

Make the modelling decision before importing thousands of records. A suitable field definition includes its data type, example value, ownership, required status and where it must appear on a product page or filter.

The official CSV guide supports certain product metafields, but **variant metafields are not imported through the standard product CSV**. Those need a suitable supported workflow instead. Do not promise that one spreadsheet carries every custom attribute across.

Optional fields should also be tested when empty. A missing value must not leave misleading labels or empty sections visible in the storefront.

## Rebuild categories, collections and filters deliberately

A source category tree is not necessarily a Shopify collection structure. Navigation, commercial groupings and searchable attributes serve different purposes even when they share similar names.

Decide which categories become collections, which source attributes become filters, and which editorial pages need to remain independent. Consider whether collections are managed manually or by rules. Verify a sample of products in each intended collection and check that filtering actually returns the expected combinations.

Old category URLs deserve their own destination mapping and redirect tests. That is an SEO workstream connected to product migration, not evidence that the data import itself succeeded.

## Verify media, prices and availability at the point of purchase

A useful pilot includes a simple product, a product with several options, variant-specific images, an unavailable variant, an item missing a secondary image and an item stocked at more than one location.

Check the selected variant on the product page and in the cart. Compare its exact price, option name, image and availability against the source record. A product that displays correctly but selects the wrong SKU when added to the cart is not a passing migration.

For multi-location stock, distinguish product data from location inventory. Shopify's [CSV guidance](https://help.shopify.com/en/manual/products/import-export/using-csv) specifies that the inventory quantity field in its sample product CSV is for single-location stores; multi-location inventory needs the relevant inventory workflow. A single quantity field is not sufficient proof that all locations were reconciled.

Also inspect image loading, accessible descriptions and responsive crops. An image URL in a CSV is not proof that the photograph is available to a real visitor.

## Run a controlled import and record exceptions

A practical migration sequence has explicit checkpoints:

1. Preserve an immutable export from the source system and the field mapping.
2. Clean malformed values, missing identifiers and duplicated SKUs.
3. Select representative products, including deliberate failure cases.
4. Import the pilot into the destination environment.
5. Compare actual values and relationships, not merely product counts.
6. Correct mapping errors and repeat the pilot until it passes.
7. Import the remaining catalogue and retain exception logs.
8. Reconcile all variants, images, prices, inventory and external updates.
9. Test actual selection and cart behaviour before approving the new store.

A destination with 1,000 visible products may still contain missing variants, products with the wrong tax setting or prices attached to the wrong option. Counts are a useful diagnostic, not the final acceptance condition.

## A small, hypothetical mapping example

Consider a garment with two colours, three sizes and stock held at two locations. The source system could describe six sellable combinations, an informational material attribute and separate quantities by warehouse.

| Source field | Intended mapping | Acceptance evidence |
| --- | --- | --- |
| SKU for each size/colour combination | Shopify purchasable variant | Correct SKU in selected variant and cart |
| Selling price | Variant price | Same price for the selected option |
| Quantity at each warehouse | Location-specific inventory workflow | Expected quantity at each relevant location |
| Material description | Structured product attribute | Correct label or filter, without an extra variant |
| Multiple photographs | Product and relevant variant media | Images load and correspond to the selection |
| Blank optional feature | Omitted or empty structured field | No misleading empty section on the page |

These are illustrative records, not a client case study or a ready-to-import CSV. Include a separate test row with a duplicated SKU and ensure it is caught before any live import.

### What should stop launch?

Block publication when a critical SKU is ambiguous, a variant can be purchased at an incorrect price, stock mismatches the authoritative system, a necessary image is missing or a required integration is unable to identify a product consistently. Record each discrepancy with source record, expected result, observed result, owner and resolution.

The [Shopify migration URL mapping guide](https://ifgdigital.com/en/ifgwebdesign/ecommerce/blog/shopify-migration-301-redirects-url-map-and-tests/) covers the related redirect work, which should be tested separately from the product import.

## Accept the catalogue only after independent checks

A catalogue is ready when the destination admin, storefront, cart and relevant connected systems agree about products and their sellable variants. This is stronger than a successful upload or a matching row count.

Before committing to a platform switch, [plan your Shopify migration with Francesco Guiducci](https://ifgdigital.com/en/ifgwebdesign/ecommerce/shopify-migration/) using an explicit source-to-destination mapping and acceptance tests.
