To keep your rankings during a website migration, record a baseline, give every old URL a planned outcome, test the new site before launch, use permanent redirects for anything that moves, and monitor Search Console, server logs and analytics afterwards. Rankings can fluctuate while Google recrawls the site. Google says a small to medium-sized site can take a few weeks for most pages to move, and larger sites take longer.
When I moved my own site, mouse-jiggler.org, from WordPress to static HTML, traffic didn't dip. Comparing the four months before and after, Search Console clicks rose from 157 to 2,163. That result needs some explaining, because the migration wasn't the only thing that changed. This guide gives you the checklist, then shows the data and what it does and doesn't prove.
What Kind of Migration Is This?
"Migration" covers very different projects. Start by naming yours, because it decides whether you need redirects at all and whether the Change of Address tool applies. The risk column is my own rating, not a Google classification.
| Migration | Risk | Redirects | Change of Address | What usually goes wrong |
|---|---|---|---|---|
| New hosting or CDN, same URLs | Lower | No | No | A firewall blocks Googlebot, or the new server is too slow |
| HTTP to HTTPS | Lower | Yes, all | No | Mixed content, or internal links and sitemaps left on HTTP |
| New CMS, same URLs | Medium | Only for URLs that disappear | No | Templates lose titles, canonicals, schema or internal links |
| Redesign, same URLs | Medium | Only for URLs that disappear | No | Useful content is cut, or key content depends on JavaScript |
| New URL structure, same domain | High | Yes, all that change | No | Unmapped URLs, chains, and old URLs left in internal links |
| New domain | High | Yes, all | Yes | Variants of the old domain are missed, or redirects are removed early |
| Merging several sites | Highest | Yes, all | Yes, for each domain | Content overlap, and too many moves at once |
The Migration Timeline
Treat these as starting points for a small to medium-sized site. A large or complex site needs more time at each stage.
4 weeks before
Record the baseline, list every URL and agree who owns each task.
1 to 2 weeks before
Finish the redirect map, test staging and write the rollback plan.
Launch day
Remove blocks, turn on redirects, test them and submit the sitemap.
Weeks 1 to 4
Check logs, errors and indexing often, and fix problems quickly.
Weeks 4 to 12
Compare equal periods by page and query. Keep redirects for at least a year.
Before Launch Checklist
Most of the protection happens here. On a live migration, the owner column is what stops a task from falling between the SEO, the developer and the site owner.
| Task | Owner | Done when |
|---|---|---|
| Export the baseline | SEO | You've saved Search Console pages and queries, GA4 landing pages and conversions. Search Console keeps 16 months, so export what you may need later |
| List every URL | SEO, developer | The list combines sitemaps, a CMS export, a crawl, server logs and Search Console's Links report, and includes images and PDFs |
| Decide each URL's outcome | SEO, owner | Every URL is marked keep, redirect or remove |
| Build the redirect map | SEO | Every moved URL has one relevant destination |
| Compare old and new templates | SEO, developer | Titles, descriptions, headings, main content, canonicals, hreflang, schema and internal links match the plan |
| Block staging safely | Developer | Staging is protected, and you have a written list of the noindex and robots.txt rules to remove at launch |
| Test the new site | Developer, SEO | A crawl is clean, and forms, checkout, downloads and mobile layouts all work |
| Check Search Console and analytics | SEO | All variants are verified, verification survives the move, and tags and conversion events fire on the new templates |
| Prepare the server | Developer, host | There's capacity for extra crawling, the firewall allows Googlebot and, if DNS changes, TTL was lowered at least a week ahead |
| Plan the rollback | Developer | The old site is kept, and someone knows how to restore it and who decides to do so |
Pick a launch date in a quieter traffic period if you have one, and avoid a date when nobody can fix things the next day. Our guides to XML sitemaps and robots.txt cover the details of those two files.
Build the Redirect Map
A redirect map is a spreadsheet with one row per old URL. It's the document the rest of the migration depends on, so it should also record the test result for each row. Download the template, then replace the example rows:
- Redirect to the closest equivalent. Google warns that sending many old URLs to one irrelevant page, such as the homepage, might be treated as a soft 404.
- Use permanent server-side redirects. Google recommends 301 or 308 where possible, and says permanent redirects don't cause a loss in PageRank. See the HTTP status codes guide.
- Skip the chains. Redirect straight to the final URL. Check that older redirects don't now lead through the new ones.
- Let dead pages die. If there's no equivalent, return 404 or 410. Not every old URL deserves a redirect.
- Know when a redirect is the wrong tool. If the old page must stay available, read canonical tag vs. 301 redirect.
Test the whole map rather than a sample. This short script reads a two-column file of old and new URLs, with no header row, and prints only the rows that don't return the expected 301. It's the check I'd run on launch day, and I tested it against live Mouse Jiggler redirects while writing this guide:
while IFS=, read -r old new; do
got=$(curl -sS -o /dev/null -w '%{http_code} %{redirect_url}' "$old")
[ "$got" = "301 $new" ] || echo "CHECK $old -> $got (expected $new)"
done < redirects.csvLaunch Day Checklist
| Task | Owner | Done when |
|---|---|---|
| Remove temporary blocks | Developer | Live pages have no stray noindex, and robots.txt is the production version |
| Turn on redirects | Developer | The full map passes the script above |
| Check key pages | SEO | Top pages return 200 with self-referencing canonicals. TechSEO Inspector shows the tags on each page |
| Crawl the live site | SEO | Internal links point to final URLs, not redirects or errors |
| Submit the sitemap | SEO | The new sitemap lists final URLs only and is submitted. Sitemap Scout can check it |
| Use Change of Address (domain moves only) | SEO | Requests are submitted for all verified variants of the old domain |
| Test real journeys | Owner | A test purchase or form reaches you, and it shows up in analytics |
After Launch Checklist
| When | Task | Owner |
|---|---|---|
| Days 1 to 7 | Check server logs for Googlebot activity and unexpected errors. Run URL Inspection on your top pages | Developer, SEO |
| Days 1 to 7 | Confirm in GA4 that traffic and conversion events are still arriving | SEO |
| Weeks 1 to 4 | Watch Page indexing for rising "Not found" and other errors. Check that sitemap URLs are being indexed | SEO |
| Weeks 1 to 4 | Update profile links, ad campaigns and your most valuable external links | Owner |
| Weeks 4 to 12 | Compare the same pages and queries with the baseline, using equal periods and last year's numbers for seasonal sites | SEO |
| At least 1 year | Keep redirects live, and longer if they still get traffic. After a domain move, Google also recommends keeping the old domain for at least a year | Developer, owner |
Case Study: Moving mouse-jiggler.org from WordPress to Static
Mouse Jiggler is my own product, so I can share the numbers. On July 3, 2026, I moved its site from WordPress with Elementor to static HTML, CSS and JavaScript. I wanted to stop paying for and maintaining WordPress, and I wanted a faster site where I controlled the code, titles and schema.
The domain stayed the same, and so did the content URLs, apart from one article that got a 301. That made it the lowest-risk kind of URL plan. A short set of server rules handled the WordPress leftovers: category, tag, author and feed URLs got 301s, and paths such as /wp-admin/ and /wp-json/ return 410.

GA4 points the same way. Comparing July 4 to November 4 with the same dates in 2025, organic users rose from 1,063 to 4,354, and organic sessions from 1,184 to 9,425. Sessions grew faster because people return to the free tool.
What I did
- URL inventory: I used Claude Code to work through the WordPress export, the old sitemap and Search Console data, then decided what to keep, redirect or remove.
- Pre-launch checks: Claude Code tested redirects, links, schema and pages before the switch.
- Rollback: the old WordPress site stayed on the server, ready to restore. I didn't need it.
- Monitoring: I watched Search Console Performance and Page indexing during the first weeks, spot-checked the 301s and 410s, looked at server logs, and confirmed that GA4 was recording traffic and events.
Nothing notable went wrong. In the week of the move, the site got 7 clicks. In the four weeks after it, it got 14, 12, 18 and 13.
How much of this was the migration?
Not all of it, and I'd be wary of anyone who claims a migration alone produces numbers like these. Here's what the data supports:
- The move caused no loss. That's what this kind of migration should do, and the weekly numbers show it.
- Existing pages improved. The 10 URLs with data in both periods went from 171 to 682 clicks. The homepage moved from position 41.4 to 18.3, and
/all-downloads/from 37.2 to 15.3. They also got rewritten titles, new schema and, in some cases, deeper content, so I can't separate those effects. - New content did most of the lifting. New URLs brought 1,510 clicks. One of them, the free online mouse jiggler launched with the static site, brought 1,250, or 58% of all clicks. I also published about 20 new pages between July and October.
- A low base flatters the percentage. The site had only 9 content pages and 157 clicks to begin with. Look at the absolute numbers. Clicks also aren't sales, and many of these visitors came for the free tool.
This case also doesn't prove much about riskier moves. The domain and URLs stayed the same, so it says little about a domain change or a URL restructure. I also changed the platform, design and titles together, against Google's one-change-at-a-time advice. On a nine-page site with a rollback ready, I accepted that risk. On a site with thousands of pages, I'd split the work up.
The useful lesson is that a clean migration protects what you have and removes technical limits. Growth came from what the new setup made easier to publish afterwards.
Mistakes That Cost Rankings
- Leaving
noindexor a robots.txt block in place. Staging rules that reach production can remove pages from Google. Check the noindex status of live templates on launch day. - Redirecting everything to the homepage. It's quick, and it throws away page-level relevance.
- Forgetting URLs that aren't pages. Images and PDFs can get search traffic and links too.
- Cutting content without a decision. A cleaner design that drops the sections answering searchers' questions can lose rankings with perfect redirects.
- Moving rendering to JavaScript. If the main content now depends on scripts, test what Google can render.
- Losing tracking. Without working analytics you can't tell a tracking gap from a traffic drop.
- Removing redirects early. Google recommends keeping them as long as possible, generally at least a year.
Planning a Migration?
Request a free website review. I'll look at a small selection of public pages and email up to five priority findings or next steps. My technical SEO audit ($255, 2-day delivery) covers eight areas, so you can fix existing problems before moving them to a new platform or check the live site after launch. If a move has already left pages out of Google, my Search Console indexing fix ($170, 2-day delivery) finds the cause and fixes it. Full migration planning can be scoped separately.
Request a free website reviewFrequently Asked Questions
Related Guides
- Understand redirects and other HTTP status codes.
- Choose between a canonical tag and a 301 redirect.
- Create and check an XML sitemap.
- Run a full SEO audit checklist.

