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)

Migrating from OpenCart to WooCommerce: SKUs, Variants and URLs

Most OpenCart migrations fail months after launch. The three things that decide the outcome are SKUs, option structures and URL redirects.

Most OpenCart to WooCommerce migrations do not fail at the import step. They fail three months later, when organic traffic has not recovered, half the product variations are missing, and nobody can match an old order to a new product.

The import itself is the easy part. What determines whether the migration works is how you handle three things before you touch the database: SKUs, option combinations, and URLs. This guide covers those three in the order they will hurt you.

Why “we can just import the CSV” is a trap

OpenCart and WooCommerce describe products in structurally different ways. OpenCart stores options and their values in separate tables and calculates combinations at runtime. WooCommerce stores variations as real child products, each with its own ID, price, stock, SKU and image.

A store with 12 products and three option types — size, colour, material — can legitimately produce several hundred variation rows in WooCommerce. The import looks fine, the front end looks fine, and then you discover that stock is being tracked at the parent level and oversells start.

So the first job is not importing. It is counting.

Phase 0: build the inventory before you build anything else

Run these counts on the existing OpenCart install and write them down. You will need them again at the end to verify the migration.

Metric Where it comes from Why it matters
Products oc_product Baseline for the final row count
Products with options oc_product_option Determines how many variation records you will generate
Option combinations Calculated, not stored The real row count after migration
Categories oc_category Hidden and disabled categories included
Manufacturers oc_manufacturer Become brands or attributes in WooCommerce
Products with specials or discounts oc_product_special Pricing rules rarely map one to one
Active customers oc_customer Decides whether customer accounts move at all
Orders oc_order Almost never migrated; document the decision

Two of these deserve emphasis. Hidden products are the most common source of “the new site is missing items” complaints — they exist in the catalogue, they are just not visible, and they usually still have URLs with traffic. Special prices are the other: OpenCart specials have date ranges and priorities that WooCommerce sale prices do not replicate directly.

SKUs: decide the rule before the first row

OpenCart has both a model field and a product_id. WooCommerce has a single SKU field whose only real requirement is that it is unique across the catalogue — including variations.

Three rules that prevent most post-migration pain:

1. Use the existing model as SKU, not the product ID. The model is what appears on invoices, packing slips and supplier correspondence. If you migrate product IDs into SKU, every future stock feed and ERP sync has to be rewritten.

2. Give every variation its own SKU. WooCommerce allows variations to inherit nothing — no SKU means no way to reconcile the variation against a warehouse record. A workable pattern is MODEL-ATTRIBUTE1-ATTRIBUTE2, generated consistently, not hand-typed.

3. Preserve the old ID somewhere. Add a custom field or a hidden attribute holding the original product_id. When someone asks why a customer’s order from last year points at the wrong product, this is the only thing that will save you.

If the source data already contains duplicate models — it often does, especially after years of manual entry — fix those before importing. WooCommerce will reject or silently skip duplicates, and you will not notice until the count check at the end.

Options versus variations

This is where most of the technical work lives.

OpenCart options have a type: select, radio, checkbox, image, input. Not all of them map to WooCommerce attributes. Select, radio and image map cleanly to a variation attribute. Checkbox usually maps to an add-on, not a variation — a product with optional gift wrapping is not twelve variations, it is one product plus an option. Free-text input has no native WooCommerce equivalent.

Sort the option types before mapping, and expect to handle them in three groups:

  • True variations — size, colour, material, capacity. Each combination is a real child product with its own stock and price.
  • Add-ons — gift wrap, engraving, extended warranty. Keep these out of the variation matrix, or you will multiply your catalogue for no reason.
  • Informational fields — free-text notes. Move to order meta or a product field.

Two further traps:

Price modifiers versus absolute prices. OpenCart stores option prices as a delta with a prefix (+ or -). WooCommerce variations store absolute prices. If a product costs 40 and the large size adds 5, the variation price is 45, not 5. Getting this wrong silently halves your revenue on those products.

Stock at option level. If OpenCart tracked stock per option combination, WooCommerce must track stock per variation. If it tracked stock only at product level, do not invent per-variation stock — you will create a warehouse mismatch that takes weeks to reconcile.

URLs: the part that decides your traffic

OpenCart with SEO URLs enabled produces paths built from the product name, often with category prefixes. WooCommerce produces a path from the product slug. They will not match, and a 404 on a product page that used to rank is a lost customer and a lost ranking.

A migration without a redirect map is not a migration, it is a relaunch.

Build the map in this order:

  1. Export every active URL from the OpenCart SEO keyword table, plus the non-SEO fallback pattern (index.php?route=product/product&product_id=123).
  2. Generate the WooCommerce URL for each corresponding product after import.
  3. Store the pairs in a two-column table — old path, new path — and keep it as a file, not only in a redirect plugin.
  4. Apply 301s at server level where possible. Plugin-level redirects work, but they run after PHP has already started, which is slower and breaks if the plugin is ever disabled.

Do not forget the non-product URLs: category pages, manufacturer pages, information pages, and the search pages that accumulated links over the years. Category redirects are usually the highest-value ones because categories attract more backlinks than individual products.

After the redirects are live, resubmit the sitemap and check that the new sitemap contains no old URLs. It is common to find the old platform still generating a sitemap from a leftover file.

What usually gets skipped, and should not

Customer accounts. Password hashes cannot be transferred between the two systems. Either force a password reset on first login, or skip accounts entirely and treat returning customers as new. Decide explicitly and write it down, because the decision affects your support load for months.

Order history. Most migrations abandon it. That is usually acceptable for B2C, less so for B2B where customers reorder from history. If orders matter, keep the old store readable on a subdomain rather than trying to force them into WooCommerce.

Reviews. OpenCart reviews have status flags and no verified-purchase link. They import, but expect to moderate duplicates and spam that was never cleaned up.

Performance. A fresh WooCommerce install with the same catalogue is not automatically faster. If the old store was slow because of 40 plugins, the new one will be slow for the same reason once the plugin list is rebuilt. Plan a performance pass after launch, not before — you need real traffic patterns to optimise against.

A verification checklist that actually catches problems

Run this after import, before launch:

  • Product count and variation count match the Phase 0 inventory
  • No duplicate SKUs; every variation has a SKU
  • Ten random products: price, stock, images, description identical to the old site
  • Every option combination that existed still exists
  • All old URLs return 301 to a live page, not to the homepage
  • Sitemap contains no old-platform URLs
  • One test order per variation-heavy product, completed end to end

That last item catches more real bugs than the other six combined. Variation-heavy products are where cart logic, stock reservation and shipping calculations all meet.

How long this actually takes

It depends almost entirely on option complexity, not product count. A 300-product catalogue with no options is a few days of work. A 60-product catalogue where every product has four option types and per-combination pricing is slower, because each product needs a mapping decision.

If you are planning a migration, the useful question is not “how many products” but “how many distinct option structures”. That number is what you should scope against. If you would rather not do that assessment yourself, we handle OpenCart migrations and the first thing we ask for is the option inventory, not the product count.

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