Free 20-point conversion & SEO audit for export stores — Free website audit — claim yours

hello@woo.one Mon–Fri · 9:00–18:00 (UTC+8)

Customers and Orders: What to Migrate and What to Leave Behind

Passwords cannot move, and most order histories should not. Deciding the reason for migrating orders first, plus the traps in subscriptions, gift cards and consent.

Products migrate. Customers and orders are a different problem: one cannot be moved intact, and the other often should not be moved at all. Both decisions are usually made by default, and the default is usually wrong.

Customers: the records move, the passwords cannot

OpenCart, Shopify, Magento and WooCommerce all hash passwords with the platform’s own implementation. There is no supported way to transfer them, and attempting to copy hashes produces accounts nobody can log into.

Three options, and you should pick one deliberately:

Option What the customer experiences When it is right
Force a password reset on first login Enters email, gets a reset link, sets a new password Most stores
Send a proactive reset email at launch Receives “we have moved, set your password” before they need it Stores with a large repeat-customer base
Do not migrate accounts Checks out as guest, no history One-time purchase catalogues, or very small stores

The second option converts best and is the one most stores skip. It costs one email campaign and removes the friction of a customer discovering, at checkout, that their account no longer exists.

Two details worth getting right: the reset email must come from a monitored, authenticated sender (see the deliverability notes in the pre-launch checklist), and the reset flow must work before you announce the move — not after customers start reporting it.

What customer data should actually move

Name, email, billing and shipping addresses, phone, marketing consent, and account creation date are the useful fields. Everything else is either platform-specific or a liability.

Be careful with consent. If your old store recorded marketing opt-ins, carry the consent with its timestamp and source. If it did not record consent, do not import those addresses into a mailing list — import them as customers who can receive transactional email only. Getting this wrong is a compliance problem, not a marketing problem.

Also apply a retention rule rather than moving everything. Accounts with no orders in three years, or with no orders at all, are usually not worth carrying. Document the rule you applied, as described in the cleanup pass.

Orders: three reasons to migrate, and one reason not to

Orders are the heaviest data and the least standardised. Before importing them, name the reason:

  • Customer service — support needs to look up past orders
  • Reporting — you need historical revenue in one system
  • Reorder — B2B customers reorder from history

If none of these is true, do not import them. A migrated order table with mismatched line-item metadata produces reports that disagree with your accounting system, and that disagreement costs more than the migration saved.

If one of them is true, decide which, because each needs different handling:

For customer service, a read-only archive is enough. Keep the old store accessible on a subdomain behind basic auth, or export orders to a simple lookup tool. Importing them into WooCommerce gets you a screen support staff will not trust anyway, because totals and taxes will not reconcile.

For reporting, export to a warehouse, not to WooCommerce. Orders in an analytics table are useful; orders in a shop database pretending to be history are not.

For reorder, import only what reorder needs: customer, product, quantity, date. Skip totals, taxes and shipping details entirely. A “past orders” list that shows product and quantity is enough for a B2B buyer to reorder from.

The trap: line items that no longer exist

Historical orders reference products and SKUs that may have been deleted during cleanup. If you import orders after deleting products, you get broken references and empty line items.

Order of operations matters:

  1. Decide the product cleanup, including what will be deleted
  2. Map historical order line items to surviving products or to archived placeholders
  3. Import products
  4. Import orders, with the mapping applied

If a product was genuinely discontinued, create it as a hidden, out-of-stock product so historical orders have something to point at. Hidden products do not pollute the catalogue and prevent the reference error — the same reasoning as keeping disposed zombie SKUs alive in the SKU section of the migration guide.

Subscriptions, gift cards and store credit

These three do not migrate cleanly anywhere, and each has a financial liability attached.

Subscriptions are usually the hardest: the payment token lives with the old gateway and cannot be moved. Plan to ask customers to re-authorise, and expect a churn spike. Communicating it well is the difference between 5% and 20% loss.

Gift cards and store credit are a liability on your books. Export balances, import them manually or via a plugin, and reconcile the total before and after. Losing a customer’s gift card balance is a legal problem in many jurisdictions, not just an inconvenience.

What to test before launch

  • A migrated customer can complete a password reset and log in
  • Their saved addresses appear and can be used at checkout
  • Their order history shows what it should show — and nothing that is wrong
  • A new customer can register without colliding with a migrated email
  • Marketing emails were not sent to non-consenting addresses

The fourth item is easy to miss: if an email exists from an old import with an empty password hash, registration can fail with a confusing error. Test it with a real migrated address.

None of this is technically difficult. It is a series of decisions that need to be made before the import rather than discovered afterwards. If you are planning a migration and want the customer and order scope assessed first, that is part of the migration work we do — and the output is a written decision, not just a database.

S

Infrastructure and security engineer at woo.one, based in Foshan, China. Works on WordPress, WooCommerce and OpenCart systems, self-hosted mail, and the operational side of running stores for exporters.

More about the team

No pressure · No sales script

Ready for a store that sells while you sleep?

Send us your current site or product list. You will receive a fixed-scope proposal, a 20-point audit and a realistic timeline within two business days.

Reply within 1 business day NDA on request 100% code ownership
Chat with us
WhatsApp Get a quote