Ecommerce Platform Migration Checklist (2026)

Most ecommerce migrations fail in predictable ways: missing 301 redirects bleed organic traffic, product data maps badly between platforms, and an integration nobody documented breaks on launch day. A migration that protects revenue is mostly preparation — a full URL map, a parallel-run window, and a tested cutover — not the data export itself. A typical mid-market replatform runs 12-20 weeks and £25,000-£120,000 depending on catalogue size and integration depth; the stores that come through clean are the ones that front-loaded discovery.

Before you start: discovery

Document everything

Your current site does more than you think. Catalogue every integration, automation, custom feature and workaround, and talk to everyone who touches the system — customer service, warehouse, finance, marketing. They know the edge cases the documentation never captured: the nightly stock file, the VAT rule for the Channel Islands, the discount that only fires for one wholesale account. Each undocumented behaviour is a launch-day incident waiting to happen.

Define what “done” means on day one

Separate must-haves from can-waits. What does the new platform have to do the moment you flip traffic, and what can ship in week two? Being explicit prevents scope creep and gives you a clean rule for the trade-offs that always surface mid-build. Write it down and get sign-off — a vague brief is how a 10-week project becomes a 20-week one.

Set realistic timelines

Migrations take longer than expected, always. Budget buffer for data cleanup, redirect testing, staff training and the inevitable surprises. A useful rule: the data export is roughly 20% of the work, and the reconciliation, redirects and integration rebuild are the other 80%. Rushing the launch to hit an arbitrary date is the single most common way a sound migration turns into a traffic loss.

The redirect map is the whole game

Poor redirect handling is the most common — and most expensive — migration failure. Every URL that changes needs a 301 redirect to its closest new equivalent. A mid-sized catalogue routinely needs 2,000-20,000 redirects once you count products, collections, blog posts, paginated pages and legacy campaign URLs. Miss a chunk and you lose the rankings those pages earned; get the mapping wrong and you confuse crawlers for months.

Crawl the existing site in full (Screaming Frog or Sitebulb), export every indexable URL, and map old to new one-to-one — never collapse everything to the homepage, which Google treats as a soft 404. Preserve canonical tags and meta data, keep slugs stable where the new platform allows, and plan for platform-specific path changes: Shopify forces /products/ and /collections/ prefixes, while BigCommerce and Magento use different category-path rules, so a like-for-like URL is often impossible and the 301 becomes mandatory. Test the full redirect map on staging before launch, then re-test the top 500 URLs by traffic the moment you cut over.

After launch, watch Google Search Console daily: a 10-30% organic dip in the first two to four weeks is normal as Google recrawls; it should recover, not keep falling. If traffic is still dropping at week six, you have broken redirects or de-indexed pages, not a normal settling period. For worked examples of clean platform moves, see our Magento to Shopify migration and Volusion to BigCommerce guides. Volusion exits in particular need SEO triage before the move — our Volusion SEO service exists for exactly that.

Data migration: budget for the cleanup

Product data rarely maps cleanly. Variants, metafields, option sets, category trees and images are structured differently on every platform, and the export almost never lands ready to use. Budget real hours for reconciliation — for a 5,000-SKU catalogue, plan on 40-120 hours of manual cleanup and validation, more if your data has accumulated a decade of inconsistencies. Always migrate a representative subset first, validate it end to end, then run the full load.

Customer and order history are usually more mechanical, but verify the things that quietly break trust: loyalty balances, active subscriptions, saved addresses and tax-exemption flags. Order history in particular has to survive intact for returns, accounting and support — losing it is a problem that resurfaces for months.

The parallel-run and cutover

Don't flip the switch overnight. Run both systems in parallel for one to two weeks: process real test orders on the new site, reconcile inventory against the old one, and let staff work in the new admin before customers ever see it. This window is where you catch the integration that times out under load and the tax rule that rounds the wrong way — before they cost you real orders. Point the 301s and DNS only once the new store is fully indexed and you have verified the redirect map in Search Console.

Post-launch: prove the metrics held

Staff your first two to four weeks heavily. Monitor conversion rate, checkout completion, error logs, Core Web Vitals and customer complaints in parallel — a migration can launch with zero downtime and still quietly lose 5% of conversions to a slower checkout or a broken discount. A new platform is also a chance to fix what was slow before; pair the cutover with our store performance guide so you launch faster, not just different. A successful migration is not “we launched without downtime” — it is “our organic traffic, conversion rate and order accuracy held or improved 90 days later.”

For a worked example of the catalogue side of this checklist, the Veluare launch case study shows a full product catalogue migrated onto Shopify with clean data: consistent naming, materials kept in titles where buyers search for them, accurate collection tagging, and no dead links on go-live day.

If you are still choosing a platform before any of this starts, our BigCommerce vs Shopify decision guide and when Shopify Plus makes sense cover the choice itself; this checklist is what you run once it is made.

Newer related guide: Shopify Ecommerce Platform Detection Methods (2026).