At a glance
- Rankings are tied to addresses: every old URL with visitors or links needs a matching target on the new website.
- Redirects are permanent (301) and go to the page with matching content, not across the board to the home page.
- Before launch, redirects, blocks, sitemap, titles and structured data are tested.
- Afterwards, several weeks of monitoring: indexing, error pages and the clicks of the most important pages.
Why a redesign costs visibility
- Old addresses lead to error pages because redirects are missing.
- Content that ranks well gets cut or dropped.
- A block from the staging environment (noindex or robots.txt) stays active.
- Internal links, titles and headings change without a plan.
- The new website loads more slowly, especially on smartphones.
Inventory before the rebuild
It starts with a complete list of all addresses of the old website. The list comes from three sources: a crawl of the website, the Search Console performance report with clicks and impressions per page, and the landing pages from web analytics. Add the targets of links from other websites, as far as they are known.
For each page, the decision is: keep, merge or remove. Pages with many clicks or links are carried over with equivalent content wherever possible.
The inventory includes a baseline: clicks and impressions per page over recent months and enquiries per channel. Only with this comparison can you tell after the redesign whether it worked.
The redirect plan
- Type
- Permanent (301 or 308), not temporary (302 or 307)
- Target
- The page with matching content, not the home page across the board
- Chains
- Straight to the target, without intermediate steps
- Removed content
- Without an equivalent: status code 410 or 404
- Variants
- With and without www, http and https, with and without a trailing slash
- Duration
- As long as possible, at least one year according to Google
Ideally, addresses stay the same. Every change to the URL structure that is not necessary creates effort and risk.
Content and internal links
Content that gets found today keeps its core statements, headings and technical terms. When two pages are merged, the answers of both pages carry over. Internal links point straight to the new addresses instead of running through redirects.
Titles and descriptions of pages with good rankings stay the same at launch wherever possible. Improvements follow once the new website has settled; that way, cause and effect can be told apart.
Testing before launch
- Check the redirect list automatically: every old address returns 301 and lands on the right target.
- No blocks: robots.txt allows crawling, no page carries noindex, canonical tags point to the new addresses.
- Generate a sitemap with all new addresses and link it in robots.txt.
- Carry over or improve titles, descriptions and structured data.
- Test forms, tracking and cookie consent.
- Measure load time on a smartphone.
The weeks after
After launch, the new sitemap is submitted in Search Console. The indexing report shows which pages were included and which errors occur; error pages with visitors get a redirect right away. Fluctuations in the first weeks are normal while Google reassigns the addresses. If the drop persists, the cause usually lies in the redirect list or in missing content.
When the domain changes, the Change of Address tool in Search Console comes into play as well.
Frequently asked questions
How long do redirects have to stay in place?
Google recommends keeping them as long as possible, but at least one year. Links from other websites to old addresses often last much longer; redirects kept permanently cost little.
Can a new design and new technology come at the same time?
Yes, but the risk grows with every simultaneous change. If addresses and content stay the same, problems are easier to trace to a cause.
How a new website gets found from the start is described in the Websites and Visibility modules.