Site Migration

Migrations are the single most common cause of self-inflicted SEO disasters. Industry surveys consistently show 30–50% of site migrations cause measurable traffic drops, and 10–20% never fully recover. The biggest wins compound, a clean migration preserves the entity equity a domain has taken years to build. The biggest losses are brutal, a bungled migration can cost six-figure monthly revenue overnight and take quarters to claw back. A methodical pre/during/post checklist is the difference between the two outcomes.

Domain migration: example.com → newbrand.com. The highest-risk category. Requires 301 redirects on every URL.

URL structure migration: /products/abc → /shop/abc. Often runs together with a CMS change.

Platform / CMS migration: WordPress → Shopify, custom → Webflow. Templates change; URLs often change too.

Protocol migration: HTTP → HTTPS. Low-risk if done right, disastrous if mixed-content errors appear.

Design / redesign: Same URLs, new HTML, different internal linking, different content depth.

Subdomain ↔ subfolder: blog.example.com ↔ example.com/blog. Both work for SEO but migrating between them must preserve links.

Internationalization: Adding or restructuring hreflang and country subdirectories.

Full URL crawl of the current site: Screaming Frog or Sitebulb. Capture every indexed URL, its traffic, and inbound links.

Map old → new URLs: A spreadsheet with current URL, new URL, status, redirect type. Every indexed URL needs a destination.

Audit backlinks: Identify the pages with the most external links. These must retain their redirects even if the rest of the site is pruned.

Benchmark baseline metrics: Rankings, GSC impressions, organic traffic, indexed pages, Core Web Vitals. You need a before picture to measure loss or gain.

Stage the new site: Crawl it on a dev domain. Fix broken links, missing canonicals, and new-structure errors before go-live.

Communicate with stakeholders: Legal, marketing, email, paid ads all depend on URL stability. Surprise migrations break everything.

301 redirects live before cutover: Every old URL 301s to its mapped new URL, including deeply buried category pages.

Update canonical tags to new URLs.

Submit new sitemap to Google Search Console.

Update robots.txt, don't block the new site, don't leave the old block in place.

Monitor 404s and redirect chains hourly for the first 24 hours.

Watch Core Web Vitals, a platform change often affects LCP and INP silently.

Keep old sitemap accessible for at least a few weeks so Google can discover the redirects.

Week 1: Crawl errors, 404s, redirect loops, sudden ranking drops. Compare GSC coverage reports daily.

Month 1: Index coverage, ranking changes, traffic delta vs baseline. Expect a temporary dip; it should stabilize.

Quarter 1: Final recovery assessment. If traffic isn't back to baseline within 8–12 weeks, something is structurally wrong.

302 instead of 301: 302 is temporary; Google doesn't pass full signals. Use 301.

Redirect chains: Old → intermediate → new burns crawl budget and bleeds signal. Redirect directly to the final URL.

Losing internal links: New templates often drop the old internal link graph. Re-audit after launch.

Forgetting sitemap updates: Google keeps crawling the old URLs until you tell it otherwise.

Skipping staging crawl: Problems found in production are 10× harder to fix.

No rollback plan: If the new site breaks, you need a tested path back to the old one, at least for the first 48 hours.

Quiet migrations: Announcing the migration to Google Search Console via the Change of Address tool (for domain moves) helps Google recognize the move.