A common misunderstanding about moving Amazon products into Shopify is that a CSV import will finish the job.
A CSV file is one way to get data into the system. It is not a migration plan.
Amazon listings are organised around ASINs, parent–child variations and marketplace browse nodes. Shopify has to assemble what the shopper sees as a product, the variants they can buy, the collections they use to navigate, the metafields that explain the item, and the inventory locations that actually fulfil orders.
If that mapping is skipped, the usual results follow: parent–child sets split into duplicate products, colour and size jammed into titles, collection pages that cannot filter, inventory that does not match, and an operations team that has to rebuild a spreadsheet every time a field changes.
This article answers one question: how to rebuild Amazon product records as a Shopify catalogue the team can operate.
Build the mapping table before you upload
Every source product should at least have these fields decided:
| Source data | Shopify target | Decide before migrating |
|---|---|---|
| ASIN / parent SKU | Product / handle | Which children belong to one buying decision |
| Seller SKU | Variant SKU | Whether it remains the inventory and system-sync key |
| Variation theme | Option / variant | Which attributes actually change the purchase |
| Bullets / A+ Content | Product description and page modules | Which lines are product facts and which must be rewritten |
| Browse node | Collection / taxonomy | How shoppers should find the product on the brand site |
| Attribute fields | Metafield | Type, unit and who will maintain the field |
| Inventory quantity | Location inventory | Which location is responsible for fulfilment |
Decide which Shopify object each field belongs to before choosing CSV import, the bulk editor, an app or the API. Use the current Shopify product CSV specification: variant columns have dependencies, and inventory across multiple locations uses a separate inventory CSV. Test updates on a small sample before importing the full catalogue.
Products and variants: define one buying decision first
A Shopify product should match one item in the shopper’s mind. A variant should match a combination of that item that can be bought directly.
If the same device comes in black or white and in two capacities, colour and capacity can be variant options. Material notes, compatible systems and installation requirements are usually not purchase options. Those belong in metafields.
For each parent–child group, answer:
- Will shoppers treat these children as the same product?
- When an option changes, do the main image, inventory, barcode or fulfilment need to change?
- Is this attribute a buying choice, or product explanation?
- Will the number of combinations make the product page hard to maintain?
Do not force every Amazon child into one Shopify product just to preserve the marketplace family. Do not split every colour into its own product either, or collection pages fill with repeated cards.
Collections: rebuild navigation around shopper intent
Amazon browse nodes serve marketplace search. They should not become the brand site’s navigation.
Shopify collection pages should help shoppers narrow the choice. Common ways to organise them include:
- by use: commuting, travel, outdoor, professional work;
- by compatibility: device model, system, size;
- by product role: starter, core, upgrade, accessory;
- by a key specification: capacity, material, power or installation method.
Write the question each collection has to answer, then choose a manual collection or an automated, condition-based collection. Conditions should rest on stable product fields, not on a hand-maintained “featured products” list. Confirm current collection types and condition options in Shopify admin before locking the rules; those details change.
Metafields: take product facts out of the copy
Many Amazon selling points should not remain a long rich-text block after they reach Shopify.
Content that is worth structuring includes:
- materials and dimensions;
- compatible models;
- power, capacity and operating conditions;
- pack contents;
- installation and care;
- certification or compliance notes;
- warranty and after-sales boundaries;
- FAQs and downloadable files.
Define the field name, type, unit, which products it applies to, and who maintains it, then import the data. Do not store every number as text. Do not store true/false values as “yes/no” strings. Do not squeeze several objects into one comma-separated sentence.
When the same content has to be reused across products — compatible devices, specification groups or material notes — evaluate Metaobjects. Do not structure every sentence for the sake of looking technical. Confirm current Metaobject capability on the store’s plan before treating it as the default.
Media: do not paste A+ Content as-is
Image migration should distinguish at least five roles:
- collection-page thumbnails;
- product main images;
- specification and detail images;
- lifestyle or in-use images;
- installation, comparison or instruction images.
Record the purpose, order, alternative text and mobile crop for each image. Dropping one long marketplace graphic into the product description makes type too small on a phone and makes it hard to update a single selling point.
A more stable split is: product facts go into fields, page structure goes into theme modules, and images only do the display work they are actually good at. Keep product identity, contents and prices accurate across Amazon and the brand site. Channel difference is not a reason to hide that two listings sell the same item.
Inventory and fulfilment: the SKU has to be a stable join key
Before moving inventory, confirm:
- every purchasable variant has a unique SKU;
- Amazon, the ERP, the 3PL and Shopify use the same SKU set;
- which system is the source of truth for stock;
- which locations serve which markets and orders;
- whether a bundle is its own inventory, or is calculated from component stock;
- how oversell is prevented when a sync fails.
A Shopify location is not only a warehouse label. It affects inventory and fulfilment. Set up locations, SKUs and fulfilment rules, then import quantities. Do not use a product title or an ASIN as the long-term inventory primary key.
Price and markets: do not just subtract marketplace fees
Brand-site pricing has to account for product role, bundles, services, currency, tax and market visibility. It is not Amazon’s selling price minus platform fees.
If different regions sell different versions, check together:
- whether the product is for sale in that market;
- currency and displayed price;
- the matching inventory location;
- language and specification units;
- after-sales and compliance copy;
- URLs and search landing pages.
Multi-market work is not a translated homepage. Products, prices, content and fulfilment have to line up. Shopify Markets features, tax settings and catalogue visibility differ by plan and change over time; verify the current feature set independently rather than assuming a previous setup still applies.
A sound import order
Work in this sequence:
- Lock the field dictionary. Sign off the mapping for products, variants, collections, metafields and locations first.
- Import a small sample. Take one simple product, one complex product and one exception.
- Check products and variants. Titles, handles, options, SKUs, barcodes and inventory policy.
- Build collections and filters. Confirm shoppers can narrow the set by a real need.
- Import structured facts. Metafields, explanations, downloads and FAQs.
- Complete media and mobile crops. Check main images, variant images and in-use images.
- Import inventory and connect fulfilment. Prove unique SKUs across locations.
- Place a full test order. Walk landing page, variant choice, add to cart, checkout and fulfilment write-back.
- Then widen the batch. Expand to the full catalogue only after the sample is stable.
Launch checklist
- Variants under the same product belong to one buying decision.
- Every variant has a unique, stable SKU.
- Collection pages follow shopper intent, not marketplace browse nodes.
- Specifications, compatibility and pack contents live in structured fields.
- Images have a clear role and a mobile crop.
- Inventory locations and fulfilment have been checked with a sample order.
- Products, prices, language, tax and stock are consistent for each market in scope.
- Old links, search landing pages and any required redirects have been planned.
- Operations can maintain core product facts without editing theme code.
Why mapping and page design have to be done together
WESWOO is a Shopify partner team that plans, designs and develops brand storefronts. It is not a tool for pushing Amazon spreadsheets into admin, and “changing platforms” is not the same as dropping the catalogue into a new theme. The work is to reorganise existing product records into a Shopify buying experience shoppers can understand, operations can edit and other systems can join.
The data model decides whether the catalogue stays controllable as it grows. Start from the relationship between products, variants, collections, metafields, Metaobjects and locations, then continue through theme architecture, editable modules, filtering and search, inventory and fulfilment connections, and Markets. Assess Shopify B2B, checkout changes and Shopify Plus against the actual business and current plan features. Shopify B2B is not Plus-only by default. Company accounts, catalogues and wholesale workflows vary by feature; checkout customisation depends on which part of checkout is being changed. Verify those boundaries independently. The point is not to stack features. It is to make data ownership, system keys, maintenance responsibility and exception handling explicit.
Design decides whether that model becomes a brand experience. The same product data still has to land on the homepage, collection pages, product pages, cart and content pages, with brand visual language, information hierarchy, a buying path and a usable mobile layout. Brand expression belongs in reusable page modules, product fields and interaction rules, so the visual system is not a one-off render, and the theme is not a generic admin shell.
When mapping and design sit in the same team, migration is less likely to stop at “the data is in, but shoppers still cannot choose and operations still cannot edit.” Validate the model, the pages and a real purchase path on a small set of representative products, then widen to the full catalogue. Split the critical steps into deliverables that can be checked, tested and accepted.
If you are preparing to extend from Amazon onto Shopify, bring the product export, variant relationships, inventory systems, target markets and current brand materials. That is enough to discuss whether the first problem is the product model, system connections, page experience or brand expression, and how WESWOO should take the next step.
Related work: Shopify storefront page design and development.
FAQ
Can Amazon parent–child listings map directly to Shopify products and variants?
Not by default. First decide whether those children belong to one buying choice, then merge, split or regroup them.
Does a product CSV solve the whole migration?
No. It can carry some product and variant data. Collection rules, structured fields, inventory locations, media, markets and system sync still need their own design and checks. Confirm which columns the current Shopify CSV supports before treating the file as a complete catalogue.
Do many SKUs mean the store needs Shopify Plus?
No. SKU count alone is not a Plus requirement. Review Shopify Plus against work you actually need: company-account pricing, catalogues, multi-market rules, checkout changes and system coordination. Treat those as reasons to compare current plan features. Shopify B2B is not Plus-only by default; catalogue limits, company-account tools and checkout customisation differ by feature and change over time, so verify them independently.