TL;DR: The checklist I use before a store switches ecommerce platforms has six phases: document business requirements, audit the current site, map every data type, build a one-to-one redirect plan, validate in staging, and monitor hard for 30 days after launch. Skip any phase and you will spend weeks chasing problems that careful pre-work would have caught in an afternoon.
Why I Built This Checklist Around Hard Lessons
The checklist I use before a store switches ecommerce platforms came directly from a migration I handled early in my ecommerce career where I genuinely thought the job was mostly technical: export the products, import them, flip the DNS, done. What actually happened was a week of broken checkout flows, missing order history for customers calling support, and a crawl of the new site that turned up hundreds of unredirected URLs that had been earning real organic traffic. That experience changed how I think about every replatforming project.
Replatforming is not a technical event with a business side effect. It is a business event with a technical execution. The right platform does not automatically fix poor conversion, weak copy, or a confusing catalog structure. What it can do is stop getting in the way of what already works, and sometimes unlock capabilities that were simply unavailable before. But only if the migration preserves what was working and nothing important falls through the gaps.
The two questions I ask at the start of every migration conversation: what is breaking badly enough that we are willing to absorb the cost and risk of moving, and what would we lose if something went wrong at launch? Those answers shape every decision that follows.
How the Checklist I Use Before a Store Switches Ecommerce Platforms Starts: Business Requirements and TCO
Before anyone evaluates a new platform, I want a written list of capabilities the store genuinely needs. That means product complexity: how many variants, whether the catalog uses subscription products, whether there are customer groups with different pricing, and what sales channels need to stay connected. Without this list, platform demos become beauty contests decided by whoever has the slickest presentation.
Total cost of ownership is the second conversation. Platform licensing is the visible number. The less visible numbers are development and customization, integration rebuild costs, data migration work, testing time, training for the team, and ongoing support contracts. I have seen stores choose a platform that looked cheaper on the monthly invoice and then spend two or three times that on integrations in the first year. Getting realistic estimates for each cost category, before signing any contract, is not optional.
Also worth documenting before the platform decision: ownership and exit terms. Can you export your own data cleanly? Who owns the customer data? What happens to the domain and store history if you stop paying? These questions are easier to ask before you sign than after.
Auditing the Current Site Before Switching Ecommerce Platforms
A full crawl of the current site, run before migration work begins, is the foundation of everything that comes later. I use Screaming Frog SEO Spider to export every URL with its status code, canonical tag, indexability flag, title, meta description, heading structure, and internal links. That output becomes the reference document the entire migration is checked against.
Alongside the crawl, I pull organic landing page data from Google Analytics 4 using the Engagement > Landing page report filtered to organic session medium, and pull revenue figures from an Exploration with Landing page as the primary dimension and Purchase revenue as the metric. I then cross-reference with Google Search Console to see which URLs are receiving clicks and impressions. These benchmarks exist for one reason: so that six weeks after launch I can compare actual performance against a known baseline instead of arguing about whether traffic dropped.
Structured data is part of the audit too. If the current site has Product, Offer, Review, or BreadcrumbList schema, I document exactly what is live and confirm the new platform will replicate it. The Product type vocabulary at schema.org covers the fields most ecommerce platforms should be marking up. Losing existing structured data at launch is an invisible but real hit to AI search visibility and to rich result eligibility.
Conversion Catalyst: Re-implementing Product and Offer structured data on the new platform before launch, rather than adding it as a post-launch cleanup task, keeps rich results active through the transition window. Treat schema markup as a pre-launch deliverable on the same checklist as redirect validation, not a cleanup item queued for the week after go-live.
Data Migration: Mapping Products, Customers, and Orders
No migration should touch production data without at least one complete test run first. I build a field-by-field mapping document that lists every data type on the legacy platform and its destination field on the new one: products, variants, SKUs, prices, inventory levels, images, alt text, collections or categories, and any custom attributes the catalog relies on. This level of detail is why the checklist I use before a store switches ecommerce platforms treats data mapping as a separate phase from the technical migration itself.
Customer data needs its own section of the mapping document. Accounts, shipping addresses, order history, loyalty points, gift card balances, subscription status, and permission levels all need to migrate and be validated with real test accounts, not sample data. Order history matters more than many teams expect. Customers who contact support and discover their order records are gone get nervous, and that nervousness turns into disputes.
The test migration produces a data reconciliation report: a row count comparison showing how many records went in versus how many came out, with discrepancies flagged before the production migration runs. I check that report line by line. It is tedious and it is worth every minute.
What the Checklist I Use Before a Store Switches Ecommerce Platforms Does With SEO Redirects
A one-to-one 301 redirect map for every indexed URL is non-negotiable. That includes product pages, collection or category pages, blog posts that bring in organic traffic, and any URL that appears in Google Search Console with impressions or clicks. Google Search Central’s documentation on redirects and site moves is explicit that 301 redirects are the right signal for a permanent URL change and that they pass link equity to the destination URL.
Building the redirect map is a spreadsheet job: old URL in column A, new URL in column B, HTTP status after implementation in column C. That column C gets filled in staging before cutover, using a redirect crawl. Any redirect that returns a 302 instead of a 301, or that chains through more than one hop, gets corrected before launch day.
After go-live, I monitor 404 errors in Google Search Console’s Page Indexing report daily for the first two weeks. A spike in 404s is the fastest signal that something in the redirect map was missed or broke during deployment. I also watch the crawl stats report to confirm Googlebot is following the redirects and indexing the new URLs rather than repeatedly crawling the old ones.
How the Checklist I Use Closes Out: Staging, Launch Day, and the First 30 Days
Staging is where every assumption gets tested. I run end-to-end purchase flows for every product type the store sells: simple products, variants, bundles, subscriptions, digital goods. Each flow is tested on mobile and desktop, across the payment methods the store actually uses, and for at least two shipping destinations if the store sells internationally. Any checkout error in staging is a priority fix before anyone discusses a launch date.
On launch day, the cutover sequence matters as much as the staging tests. I work from a written go-live checklist: verify DNS propagation, confirm the SSL certificate is active on the new platform, run a live purchase on the new store before removing the maintenance page, and check that order confirmation emails are firing correctly. The first real transaction on the new store is the most important QA step you can run, and it should happen before you announce the cutover publicly.
The first 30 days after launch are a monitoring phase, not a celebration phase. I track four metrics daily for at least the first two weeks: revenue and conversion rate against the pre-migration baseline, unresolved crawl errors flagged in Search Console, integration error logs for the payment gateway and email platform, and customer support tickets that mention an order or account problem. After the two-week mark, weekly reviews replace daily checks until performance has fully stabilized.
Platform migrations are not won at launch. They are won in the weeks of preparation that happen before a single file is moved. Every phase of this checklist, from documenting requirements through the 30-day monitoring window, represents a category of problem I have either fixed the hard way or watched someone else fix. The store that arrives at go-live with clean redirects, validated data, and tested checkout flows is in a fundamentally different position than one that treated the technical cutover as the finish line.
Quick Takeaways
- Document business requirements and total cost of ownership before evaluating any platform, not during demos.
- Crawl the current site and benchmark revenue by landing page before migration begins, so you have a real baseline to compare against after launch.
- Run at least one complete test migration and produce a data reconciliation report before touching production data.
- Build a one-to-one 301 redirect map for every indexed URL and validate every redirect in staging before cutover.
- Monitor revenue, conversion rate, 404 errors, and integration logs daily for the first two weeks after go-live.
Frequently Asked Questions
- How do I know when it is time to switch ecommerce platforms?
- The clearest signal is when the platform blocks capabilities the business genuinely needs, such as handling product variants, subscriptions, or multiple sales channels, and the workarounds required are consuming more time and money than a migration would. If the platform limits growth rather than enables it, and that pattern holds across several quarters, the conversation is worth having seriously.
- What data is most likely to get lost during an ecommerce platform migration?
- Customer order history, loyalty balances, gift card values, and subscription status are the data types most commonly lost or corrupted during a migration. These rarely appear in basic import templates and often need custom field mapping to transfer correctly. A thorough pre-migration field-mapping document and a test migration with a data reconciliation report catch most of these gaps before they reach customers.
- How long does an ecommerce platform migration typically take?
- A realistic timeline for a mid-sized store with custom integrations runs from three to six months, covering requirements documentation, platform selection, data mapping, development, test migrations, staging validation, and a controlled launch. Stores with simpler catalogs and fewer integrations can move faster, but compressing the testing phase consistently leads to post-launch problems that take longer to fix than the time saved.
- What is the biggest SEO risk when switching ecommerce platforms?
- Missing or incorrect 301 redirects are the most common and most damaging SEO issue in a platform migration. When indexed URLs return 404 errors after launch, the organic traffic and link equity those pages had earned are effectively lost. Building a complete redirect map before launch, testing every redirect in staging, and monitoring for 404 spikes in Google Search Console immediately after go-live are the three steps that prevent most of that damage.
- How do I validate a new ecommerce platform before going live?
- Validation happens in a full staging environment that mirrors production as closely as possible. Running complete purchase flows for every product type, payment method, and shipping destination the store uses, then checking for errors in checkout logs and email triggers, surfaces problems that would otherwise appear in front of real customers. No launch date should be set until every critical checkout path passes cleanly on both mobile and desktop.

