On this page
Before the move: crawl and benchmark
Crawl the existing site and export every URL with its title, status, canonical and indexability. Pull the last 12 months of clicks and impressions by page from Search Console and list the pages with the most backlinks. These are the pages whose loss would hurt.
Decide what is changing: domain, URL structure, platform, design, or content. The more that changes at once, the harder a drop is to diagnose, so where you can, split the work into separate launches.
Take a full backup and keep the old site available, read-only, for a few weeks after launch. Save screenshots of key pages and a copy of the old robots.txt and sitemap. If something goes wrong, having the old version to compare with turns a guess into a diagnosis.
Map and implement the redirects
Build a spreadsheet with one row per old URL and its new URL. Match each page to the most relevant new page, not to the homepage. Use permanent server-side redirects, 301 or 308, and point each straight to the final destination, avoiding chains. Google treats redirects as a strong signal that a page has moved.
Pages that no longer have an equivalent can return a 404 or 410, but only after removing internal links to them. Keep the redirects live for at least a year, longer if the old URLs have backlinks. Fix broken links helps catch what slips through.
Test the map with a crawler in list mode: feed it the old URLs and check that each one returns a single 301 to the expected page with a 200. Look for chains, loops and redirects that land on a 404. Pay extra attention to URLs with parameters, trailing slashes and uppercase letters, which often fall through the rules.
Use patterns where you can. If every product moved from one folder to another, one rule can cover thousands of URLs, but test it on a sample first. Hand-map the high-value pages, such as the top fifty by traffic and by backlinks, and let rules handle the long tail. Keep the map in version control or a shared sheet so launch-day questions have one answer.
Want the articles written for you?
Researches keywords and what competitors rank for. Publishes to WordPress, Webflow, Shopify or Framer once you approve.
Test on staging, then launch
Block the staging site from search engines with a login or noindex, and test the redirect map against the old URL list. On launch day, remove the staging block. The classic migration disaster is a noindex or a robots.txt disallow carried from staging to production. Check both first.
Update the new site’s canonical tags, internal links, hreflang if you use it, and the XML sitemap to the new URLs. Keep the old sitemap briefly so Google can find the redirects. If the domain changes, use the Change of Address tool in Search Console for the old property.
Check the live site’s basics against the old one: titles, meta descriptions, headings, structured data, hreflang and internal links. A redesign often drops these without anyone noticing, and a missing H1 on every product page is a sitewide change with a sitewide effect.
Different migrations carry different risk
A domain change is the riskiest, because every URL and the brand’s history move at once. Switching from HTTP to HTTPS is lower risk but still needs redirects. A platform change, say to Shopify or WordPress, usually changes URL patterns and templates. A redesign on the same URLs is the safest, but can still break titles, headings and internal links.
Match the care to the risk. For a redesign, a crawl comparison and a title and heading check may be enough. For a domain change, plan the redirect map, the Change of Address, the sitemap and the monitoring in detail.
The mistakes that cost the most
Redirecting everything to the homepage. Forgetting to remove a noindex from staging. Changing URLs for no reason. Dropping pages that earned links without a redirect. Launching on a Friday with nobody watching. Making content changes at the same time, so the cause of any drop cannot be separated.
If traffic does fall, the traffic drop guide gives the order to check things in. Most migration problems can be reversed quickly if you find them in the first days.
After launch: verify and monitor
Crawl the live site, test samples with the URL Inspection tool, and watch the Pages report for spikes in “Not found” and “Excluded by noindex.” Compare daily clicks to the benchmark and expect some fluctuation in the first weeks as Google recrawls.
Run the technical SEO checklist against the new site. Sprites can repeat site crawls and index health checks on a schedule and read Search Console after launch, flagging errors for you to approve fixes.
Build a short daily check for the first two weeks: indexed page count, crawl errors, clicks and impressions for the top fifty pages, and a spot check of five key conversions. Set an alert threshold, for example a fall of twenty percent against the benchmark, and agree who investigates. Fast detection is what keeps a fixable error from becoming a lost quarter.
Questions
Will a website migration hurt my SEO?
It can cause a temporary dip, but a well-planned migration with correct redirects usually recovers within weeks. Big losses almost always come from missing redirects, blocked pages or major content changes.
Should I use a 301 or a 302 redirect?
Use a permanent redirect, 301 or 308, for a migration. A 302 or 307 signals a temporary move.
How long should I keep redirects?
At least a year, and indefinitely for URLs that still have backlinks or bookmarks. Removing them early loses link value.
Do I need to tell Google about the move?
For a domain change, use the Change of Address tool in Search Console and verify both properties. For URL changes on the same domain, redirects and a new sitemap are enough.
How long until traffic recovers?
Often a few weeks for a clean move, and longer for large sites or when content also changed. If nothing recovers after a couple of months, audit redirects and indexing.