How to Redesign a Website Without Losing Google Rankings

Plenty of companies stay on a website they dislike because of one fear, and it is a reasonable one: that a redesign will destroy the search traffic they spent years earning. It happens. It is also entirely preventable.

Traffic is almost never lost because the design changed. It is lost because URLs changed and nothing told Google where they went. Here is the sequence that prevents it.

Before anything is built: inventory the old site

You cannot preserve what you have not listed. Collect every URL that currently exists — from the sitemap, from analytics, from Search Console, and by crawling the site. Not just the pages someone remembers: old blog posts, landing pages from campaigns, PDFs, category and tag archives.

Sort them by the traffic and links they actually receive. This list decides what the new site must account for, and it is the single most important artefact of the whole project.

Map every old URL to a destination

Each URL on that list gets one of three fates, decided deliberately:

• Stays the same. The safest outcome. If a URL works and ranks, changing it should require a reason.
• Redirects to the closest equivalent. Permanent redirect, pointing at the page that best serves the same intent — not the homepage. A redirect to the homepage tells Google the content is gone, and it is treated roughly like a dead page.
• Is deliberately removed. Genuinely obsolete content returns a clean “gone” response so it leaves the index quickly instead of lingering.

The mistake that costs the most is the lazy fallback: everything that does not match redirecting to the homepage. It looks tidy and it throws away the value of every one of those pages.

Validate the map before launch, not after

Every redirect should be tested for two things: that it lands on a page returning a success response, and that it gets there in one hop. Chains — old URL to intermediate URL to final URL — dilute the signal and slow crawling, and they appear easily when a site has been migrated more than once.

This is mechanical work and it is where redesigns are actually won or lost. It should be a checklist someone signs off on, with the result recorded.

Carry the invisible parts across

Design gets the attention; these get forgotten:

• Titles and meta descriptions — curated per page, not regenerated from a template.
• Structured data — the markup that produces rich results.
• Canonical tags — pointing at the new addresses.
• The www / non-www and HTTPS decision — one canonical version, everything else redirecting to it. Sites competing with themselves across variants is a common and quiet loss.
• Analytics and conversion tracking — reconfigured and verified. It is remarkably common for a new site to launch without recording form submissions at all, so nobody notices the drop in enquiries for weeks.

The first month after launch

Submit the new sitemap immediately and watch indexing coverage. Expect a short period of fluctuation — Google has to recrawl and reprocess, and positions can move for a couple of weeks. That is normal and not a reason to panic or revert.

What is not normal: pages reported as not found, redirect errors, or a sustained decline after a month. Those are signals to act on, and catching them in week one is very different from discovering them in the next quarterly report.

We do this on our own site too

This site was consolidated from two separate installations into one. Every legacy URL was inventoried, mapped to its equivalent, and validated to resolve in a single hop to a live page before anything was switched. Obsolete pages were deliberately retired rather than redirected to the homepage.

We describe it as a method rather than a promise because we run it on our own property. If you are weighing a rebuild, see our approach to website and e-commerce development — or tell us the URL and we will tell you honestly whether a rebuild is worth it at all.