A Shopify migration is not a single database copy. Product records, customers and historical orders can usually be imported when the source data is clean and mapped to Shopify’s schema. Passwords, platform-specific functionality, complex promotions, editorial layouts, app behaviour, analytics continuity and many SEO signals do not transfer automatically. They must be redesigned, rebuilt, redirected or revalidated.

The safest programme separates portable records from platform behaviour. It inventories every data object, defines the target owner and acceptance rule, tests representative samples, performs a controlled delta migration and validates the live store before traffic is switched. A successful migration preserves commercial continuity—not merely row counts.

Planning a high-risk Shopify migration? Project Supply can assess your source platform, target architecture and cutover risks before implementation: Project Supply Ecommerce Development

The four migration layers

1. Structured commerce records

Products, variants, customers, orders, inventory and gift-card records are structured objects. They can often be exported, transformed and imported, but field compatibility matters. A source platform may support product models, price lists, customer attributes or order states that do not have a direct Shopify equivalent. “Imported” therefore does not necessarily mean “operationally equivalent.”

2. Content and presentation

Pages, blog posts, navigation, media and product copy may be exportable as text or files, yet their page structure belongs to the old theme or content model. Rich layouts, reusable blocks, embedded widgets, forms and responsive behaviour frequently need reconstruction in the new Shopify theme.

3. Business behaviour

Discount logic, subscriptions, loyalty rules, B2B workflows, product personalisation, tax logic, fulfilment routing and returns are behaviours, not simple records. They must be mapped to Shopify native features, apps, Functions, custom applications or changed operating processes.

4. Discovery and measurement

URLs, redirects, canonical tags, structured data, consent state, analytics identifiers, advertising audiences and attribution conventions determine whether customers and search engines experience continuity. These assets require their own migration workstream.

What usually transfers cleanly

Products and basic variants

A well-formed product catalogue normally transfers most reliably. Typical portable fields include title, description, vendor, product type, tags, price, SKU, barcode, weight, inventory quantity and image references. Variant option names and combinations can also transfer when they fit Shopify’s current product model.

Before import, normalise duplicate SKUs, invalid characters, inconsistent units, missing option values and orphaned images. Decide whether legacy categories become Shopify product categories, product types, tags, metafields or collections. This classification decision affects navigation, filtering, feeds and reporting after launch.

Customer profiles

Names, email addresses, phone numbers, addresses, tags and approved custom attributes can usually be transformed into Shopify customer records. Consent fields require special care: marketing permission should not be inferred merely because an email address exists. Preserve the source, timestamp and legal basis available in the legacy system.

Customer passwords generally cannot be migrated in a reusable form because platforms store credentials differently and should not expose password hashes. Plan account activation, password reset or passwordless access as a customer-experience workstream. Measure activation completion and support demand after launch.

Historical orders

Order history is valuable for service, returns, lifetime-value analysis and finance reconciliation. It can often be migrated through a specialist tool, app or API, but the target representation may not behave like a native new Shopify order. Confirm whether migrated orders affect inventory, customer totals, tax reports, fulfilment status, returns and notifications.

Do not import the entire order archive before testing a small set that includes refunds, partial fulfilments, discounts, multiple currencies, taxes, gift cards and cancelled orders. Finance and customer-support teams should approve the resulting representation.

Images and media

Product images are usually portable when the source URLs are accessible and the files meet Shopify requirements. Download and retain a source archive before migration. Verify image order, alt text, variant association, aspect ratio and filename quality. Embedded video, PDFs, 3D files and externally hosted assets may require separate handling.

What transfers only after transformation

Collections, categories and filters

Legacy category trees rarely map one-to-one. Shopify collections can be manual or rule-driven, while storefront filtering depends on product attributes and configuration. Rebuild the information architecture around customer discovery rather than reproducing every historical category.

Custom fields and metafields

Custom attributes can often be preserved through Shopify metafields or metaobjects, but only after defining namespaces, data types, validation and ownership. Free-text migration is easy; a governed model that supports templates, search, integrations and future editing requires design.

Inventory and locations

Inventory quantities can transfer, but location identity, committed stock, incoming stock, safety buffers and fulfilment rules may not. Freeze or reconcile inventory close to cutover, map every legacy warehouse to a Shopify location and specify which system becomes authoritative after launch.

Discounts, gift cards and store credit

Simple promotion codes may be reproducible. Complex stacking, customer eligibility, bundles, tiered pricing and platform-specific promotion engines require redesign. Gift-card balances and liabilities must be reconciled with finance; do not assume that importing a code recreates its security and accounting behaviour.

Content pages and blog posts

Article text, titles, dates and authors may be portable, but internal links, embeds, tables, schema, images and layout blocks need validation. Preserve content only when it serves current search or customer intent. Migration is an opportunity to consolidate thin or duplicated pages rather than copying technical debt.

What normally must be rebuilt

Theme, templates and frontend components

The source theme cannot be copied into Shopify as a functioning theme. Rebuild product, collection, cart, account, search, editorial and campaign templates against Shopify’s theme architecture or a deliberately chosen headless storefront. Validate accessibility, responsive behaviour, performance and editor usability.

Apps and integrations

Connections to ERP, WMS, PIM, CRM, help desk, subscriptions, reviews, loyalty, search, payments and marketing tools must be reconnected and retested. An integration inventory should record owner, authentication method, data direction, frequency, failure handling, privacy classification and cutover dependency.

Checkout and payment behaviour

Payment credentials, stored cards and platform-specific checkout customisations do not simply move. Confirm gateway eligibility, supported markets, accelerated payment methods, subscription token handling, fraud workflows and refund processes with the relevant providers. Never promise that payment tokens can transfer without explicit provider support.

Automations and operational workflows

Emails, webhooks, fulfilment rules, fraud review, order tagging, returns, customer-service macros and finance exports must be recreated. Map each workflow from trigger to action, owner and exception path. Remove automations that compensate for limitations of the old platform.

SEO assets that require deliberate migration

Create a URL inventory from the source sitemap, analytics landing pages, Search Console data, backlink reports and internal crawl. Map each valuable legacy URL to the closest relevant Shopify destination. Use permanent redirects for true replacements, preserve strong URLs where practical and avoid redirecting unrelated pages to the homepage.

Rebuild titles, descriptions, canonical tags, robots directives, hreflang where relevant, image alt text and structured data. Update internal links so users and crawlers reach final URLs without unnecessary redirect hops. Submit the new sitemap only after canonical production URLs are confirmed.

Track indexed pages, excluded pages, 404s, impressions, clicks and ranking groups before and after launch. A short-term fluctuation can occur, but unbounded loss usually indicates mapping, rendering, canonicalisation or discovery problems that require investigation.

Need a migration plan that protects organic demand as well as store operations? Review the implementation path with Project Supply: Talk to Project Supply

Migration delivery framework

Phase 1: discovery and inventory

List data objects, fields, record counts, source quality issues, URLs, integrations, workflows, reports and compliance constraints. Assign an owner and acceptance criterion to every object. Capture baselines for revenue, conversion, page speed, indexed URLs, organic landing pages, order volume and support contacts.

Phase 2: target design

Define the Shopify product model, customer model, locations, collections, metafields, app stack, integration boundaries, theme architecture and reporting model. Record every legacy capability as keep, replace, redesign, retire or defer.

Phase 3: pilot migration

Import representative samples rather than convenient samples. Include the largest product, the most complex variants, customers with multiple addresses, international orders, refunds, subscriptions and problematic characters. Verify with business users, not only developers.

Phase 4: rehearsal

Run a full-volume rehearsal in a controlled environment. Measure extraction time, transformation failures, import duration, API limits, manual corrections and reconciliation differences. Practise rollback and document the final freeze window.

Phase 5: cutover

Freeze the required source writes, take backups, run the final full or delta migration, reconcile counts and values, switch integrations, validate checkout and fulfilment, update DNS where necessary and activate redirects. Maintain a named command owner and an incident channel.

Phase 6: hypercare

Monitor checkout, payments, inventory, fulfilment, customer access, search visibility, analytics and support. Triage problems by commercial severity. Keep the legacy environment available in read-only form for the agreed retention period, subject to security and licensing constraints.

Acceptance criteria that matter

Use more than record counts. Products should render correctly, variants should select and price correctly, inventory should reconcile by location, customers should access accounts through the planned route, historical orders should support service tasks, payments should settle, tax and shipping rules should behave as approved, and integrations should recover from errors.

For content, validate high-value URLs, redirects, canonical tags, structured data, navigation, onsite search and internal links. For analytics, verify purchase events, revenue, currency, identifiers, consent behaviour and advertising destinations using test orders.

Common migration mistakes

Teams commonly migrate dirty data without ownership, copy obsolete content, discover subscription constraints too late, underestimate app replacement, treat redirect work as a launch-day task, skip full-volume rehearsal, allow both systems to remain writable during cutover, and declare success when the new homepage loads.

The corrective rule is simple: every migrated object needs an owner, transformation rule, acceptance test, reconciliation method and fallback. If any of those is absent, the item is not migration-ready.

Commercial decision model

Estimate migration cost across discovery, data remediation, implementation, integrations, theme work, licences, SEO, testing, training, cutover and hypercare. Compare this with the cost of retaining the existing platform, but also price the risks of downtime, ranking loss, manual operations and delayed growth.

A fixed launch date should not override readiness gates. Approve cutover only when critical data, checkout, payment, fulfilment, support, analytics and redirect tests pass. Defer non-critical enhancements instead of compromising the commercial core.

For hands-on data mapping, Shopify architecture and cutover support, contact Project Supply: Talk to Project Supply