Shopify migration without chaos - redirects, data, and day-one revenue

You preserve SEO in a Shopify migration by crawling every indexed URL before rebuild, mapping one-to-one 301 redirects to the closest Shopify equivalent, validating redirects in staging, and monitoring Search Console after cutover. Brands that treat migration as a visual refresh launch into broken equity: 404s, missing SKUs, unpaid apps, and paid traffic landing on dead paths.

Propeller runs migrations as commerce architecture from Santa Monica: parallel environments, staged cutover, and a post-launch window for performance and CRO. If Plus is on the table, read Shopify Plus vs Shopify before you conflate plan upgrade with URL migration. If collection equity matters, plan taxonomy with collection SEO architecture.

Redirect map discipline (non-negotiable)

Inventory every URL that earns: collections, PDPs, blog, landing pages, campaign paths. Shopify’s URL prefixes (`/products/`, `/collections/`, `/blogs/`) rarely match Magento, WooCommerce, or BigCommerce paths — so the redirect map is the spine of the project. Never bulk-redirect leftovers to the homepage; Google treats irrelevant redirects as soft 404s. Collapse chains. Handle reserved paths and query-string URLs at the edge when Shopify cannot.

• Source crawl: Screaming Frog / site crawl + Search Console top URLs + paid landing inventory.
• Map columns: legacy URL → Shopify URL → status (301) → owner → QA date.
• Rules: one-to-one preferred; category→category; never orphan high-equity PDPs into a collection dump.
• Chains: A→B→C becomes A→C before launch.
• Leftovers: intentional 410 or curated closest match — not homepage hope.
• Edge cases: vanity campaign paths, uppercase collisions, trailing slash policy documented once.

Data, apps, and ops as one cutover

Migrate products, customers, and order history with reconciliation — not hope. Decide which apps earn their keep on the new stack. Train staff on admin and fulfillment before you flip DNS. Day-one revenue is a planning metric.

• Catalog: variants, metafields, images, inventory locations — sample audit 100 SKUs by hand.
• Customers: tags, consents, and marketing status preserved for Klaviyo.
• Orders: open orders and history depth agreed with finance and CS before cutover.
• Apps: kill zombies; prefer native Plus/Shopify features where they retire paid rent.
• Payments & tax: real transactions in staging; Shop Pay and express methods verified.
• 3PL / OMS: inventory truth and fulfillment routing tested with a dry-run order.

SEO and analytics after the flip

Redirects are necessary but not sufficient. Preserve titles and canonical strategy on collections and PDPs. Re-submit sitemaps. Watch coverage and 404 reports daily for two weeks. Confirm purchase events still reach Meta and Google — broken measurement after migration looks like a “channel failure” when it is a tag failure. Tie this to Meta CAPI if paid is material.

Cutover timeline that respects revenue

• T-30: freeze net-new URL inventiveness; finalize redirect map; app decisions locked.
• T-14: full staging checkout, mobile, and fulfillment dry-run; staff training recorded.
• T-7: DNS and SSL plan; rollback criteria written; on-call roster named.
• T-0: flip, smoke test money paths, spot-check top 50 redirects, pause risky campaigns if needed.
• T+14: Search Console triage, CRO quick wins, app cost reconciliation.

Content, redirects, and paid paths as one equity set

Brand teams often migrate “the store” and forget campaign landers, influencer vanity URLs, and PDF-linked SKUs from wholesale one-sheets. Those paths still appear in ad accounts and sales decks on Monday after launch. Pull them into the same redirect workbook. If a campaign URL cannot map cleanly, retire the campaign before cutover or rebuild the lander on Shopify with the old path redirected.

• Export active URLs from Meta, Google Ads, email templates, and QR codes — not only the sitemap.
• Preserve UTM conventions so analytics comparisons survive the platform change.
• Re-point product feeds and shopping channels after catalog IDs change.
• Update customer-facing bookmarks in help center and packaging inserts on a dated checklist.

Parallel environments and rollback criteria

Chaos shrinks when rollback is boring and written. Keep the legacy store readable temporarily if DNS strategy allows. Define what triggers a pause: checkout failure rate, payment declines beyond baseline, inventory desync, or redirect error spikes. Name the human who can call the pause without a committee. Migrations fail quietly when nobody owns the red button.

Propeller stages Shopify (or Plus) in parallel, rehearses cutover with the same checklist twice, and keeps a post-launch CRO window so the new theme is not frozen as “done.” Architecture first, aesthetics second, hero videos never on the critical path.

Platform-specific traps worth naming

Magento and Adobe Commerce migrations often hide layered navigation URLs and custom routers that never appear in a naive crawl. WooCommerce migrations leak equity through mixed http/https, www variants, and plugin-generated endpoints. BigCommerce and Salesforce Commerce Cloud bring their own path conventions and promotion engines that do not map 1:1 to Shopify discounts. Inventory the weird URLs early — the “happy path” catalog is never the whole story.

• Magento: layered nav, custom routers, store-code prefixes, and legacy .html PDPs.
• WooCommerce: plain permalinks vs shop base, product category bases, and leftover page builders.
• BigCommerce: custom web paths and residual blog structures.
• SFCC: locale pathing and campaign tech that marketing still emails after cutover.
• Custom stacks: deny-list anything that 200s but is not Shopify-representable — decide 410 vs rebuild.

Propeller has migrated these stacks into Shopify and Shopify Plus with redirect QA treated as a deliverable, not a courtesy. If your RFP does not mention the redirect workbook, the agency is selling a theme.

Planning a platform move? Talk Shopify migration or start a project. Bring your current platform, catalog size, and must-keep URLs — we’ll pressure-test the redirect map before anyone picks a theme.