
A website redesign is one of the most positive decisions a company can make: updated graphics, finally-clear copy, better speed, and an experience designed for mobile. Yet it is also the moment when many companies lose, within a few weeks, the visibility they built on Google over the years. Not by bad luck, but because a website redesign affects exactly the elements Google uses for rankings: page URLs, content and structure.
The good news is that this is a manageable risk. This article explains the method we use when we need to redesign a company’s website while protecting its existing rankings, from the initial assessment to the checks carried out in the following weeks.
Why a website redesign can be costly on Google
To Google, every page is a URL with a history: the searches it appears for, the links it receives from other websites, and the engagement signals accumulated over time. When that URL changes or disappears, the history does not automatically transfer to the new page. Google finds an error where there used to be a working page, and starts again from scratch.
The damage usually comes from three directions at the same time:
- URLs changed without redirects, turning ranking pages into 404 errors
- content cut or rewritten, removing exactly the text that answered users’ searches
- forgotten technical settings, such as the block on search engines left active after development
None of these things is visible when looking at the new website: the design is perfect, everything works. The problem appears in the data a few weeks later, which is why companies planning a website redesign tend to underestimate it.
Before the website redesign: the starting snapshot
The most important work is done before touching a line of the new project. You need an accurate snapshot of the current situation, because without knowing what works it is impossible to protect it.
From Google Search Console, export the pages receiving clicks over the last twelve or sixteen months, together with the related searches. With a crawling tool, obtain the complete list of existing URLs, including images and PDFs. External-link data identifies the pages receiving links from other websites.
By combining this information, you get the list of untouchable pages: those that bring traffic, contacts or links. Their content can be improved and their design completely changed, but their value must be transferred carefully.
Saving traffic data for the same period of the previous year is also essential. Two months from now, when someone asks whether a drop is caused by the new website or by the season, you will need a real comparison, not a memory.
The redirect map: the most boring and most important document
When you decide on a website redesign, the redirect map has more impact on the outcome than any graphic choice. It is a spreadsheet with two columns: on the left, every URL from the old website; on the right, the URL on the new website that inherits its role. Each row becomes a 301 redirect, meaning a permanent instruction telling browsers and search engines, “this page has moved here”.
The practical rules are few and non-negotiable:
- every old page points to the equivalent new page, the one answering the same question, never to the homepage for convenience
- redirects are direct: no chains from A to B to C
- use 301, not 302, which indicates a temporary move
- redirects stay active for a long time: Google recommends keeping them for at least one year, and in practice it is better never to remove them
The same Google guide on moving a site puts URL mapping at the centre of the process. And if important URLs can remain unchanged, even better: the safest migration is the one where the pages that matter do not change their address at all.
What to save from the content when you decide on a website redesign
A new website is often an opportunity to streamline, and that is the right thing to do. But cutting without looking at the data is the fastest way to lose rankings. Before deleting or rewriting a page, check whether it brings traffic and for which searches.
The elements that always need careful transfer are the same:
- title and meta description of every page, which do not automatically move from one platform to another
- H1 and H2 headings and the text answering the main searches
- images and PDFs, which can receive traffic from Google Images and links from other websites
- structured data and internal links, updated to the new URLs
- for multilingual websites, the language indicators connecting the translated versions
Speed also matters: a new website that is slower than the old one starts with a disadvantage that no graphic design can compensate for. We discuss it in our guide to Core Web Vitals.
The five mistakes that make traffic collapse
In failed migrations, the same mistakes appear again and again:
- Development restrictions left active. During development, protect the staging environment with restricted access. A noindex instruction prevents indexing when Google can read it; robots.txt blocks crawling and does not guarantee that a URL will stay out of results. Before launch, check that the public website is accessible to crawlers and eligible for indexing.
- All redirects pointing to the homepage. Google interprets them as missing pages and removes visibility from each one.
- The blog deleted. Articles considered old that were quietly bringing an important share of visitors.
- Changing everything on the same day. Domain, structure, content and platform all at once: when something goes wrong, it is impossible to know what caused it.
- No checks afterwards. The website is published and nobody looks at the data for weeks, until the drop becomes visible in contact requests.
The day of publication
The launch must be prepared as a technical operation, not a celebration. In the hours after going live, check that the block on search engines has been removed, that redirects respond correctly across a broad sample of old URLs, that the XML sitemap contains only the new URLs, and that it has been submitted in Search Console.
The analytics system must also be checked immediately, because interrupted measurement during the migration makes it impossible to understand what is happening. For a local business there is one more step: update the website link in the Google Business Profile and in directory listings, as we explain in our guide to local SEO.
The weeks after: what to check
Some fluctuation after a website redesign is normal: Google needs to recrawl the pages and connect the old URLs with the new ones, and on a medium-sized website this can take several weeks. What is not normal is a decline that continues.
In the following weeks, keep an eye on three things:
- 404 errors reported by Search Console: redirect to an equivalent page when one exists; for removed content without a replacement, a 404 or 410 response can be appropriate
- the untouchable pages, checked one by one for their main searches
- organic traffic, compared with the same period of the previous year and not with the previous month
For the most important external links, it is worth asking the websites hosting them to update the URL: the redirect works, but a direct link is always more solid.
Website redesign with a domain change: double the caution
If, in addition to the website redesign, you also change the domain name, the risk increases. In this case, Search Console provides the Change of Address tool, to be used after activating the redirects. The old domain should then be kept renewed for years: letting it expire means losing all the redirects in one go.
When possible, the more cautious choice is to separate the two operations. First change the domain while keeping the website identical, then, once traffic has stabilised, redesign it. It requires a little more time, but it makes it easier to understand immediately where to intervene if something goes wrong.
What to ask whoever proposes a website redesign
Many quotes carefully describe the design, pages and functions, and do not say a word about what will happen to rankings. Before entrusting someone with a website redesign, it is worth asking a few precise questions, because the answers say a lot about the working method:
- how will the current URLs be handled? The right answer talks about a page-by-page map, not a generic “we’ll add redirects”
- who checks the data before and after launch? Someone needs to monitor Search Console in the following weeks, and it should be clear who
- what happens to the content that currently brings traffic? A serious answer starts from the numbers, not from which pages look old
- will the new website be faster or slower than the current one? It should be measured, not promised
If the person proposing a website redesign cannot answer, or says that “Google will sort itself out”, the risk is yours. There is no need to distrust everyone: you need to insist that the visibility you have earned is treated as an asset to be transferred, not as a side effect to notice later.
There is also a timing aspect to consider. A properly executed migration adds a few days of work to the project, dedicated to mapping, checks and monitoring, and those days should be included in the schedule from the beginning. Cutting them to get online sooner is the most expensive saving you can make.
Where to start
A professional website, as we explain in our guide on what really makes a professional website, is not judged only by the day it goes live, but by what happens in the months afterwards. A website redesign carried out methodically means improving the image without paying the price in visibility, and reaching launch day knowing exactly what to monitor.
If you are thinking about a website redesign and want to understand first which pages need protection, request a free consultation: we start from your real data, not from a one-size-fits-all model.

