Why a website relaunch rarely fails on design
Most relaunches look good on the day they go live. The new site is faster, tidier and works better on mobile devices. The problems surface weeks later: enquiries drop, certain pages can no longer be found in search, and the sales team misses leads that used to arrive in the CRM automatically.
The design is rarely to blame. A website that has been online for a few years is more than its surface. It has addresses that other sites point to, content that has earned a settled position in search engines, forms that pass data to other systems, and tracking code against which campaigns are measured. None of this is visible in the design, which is why it is missing from so many project plans.
A relaunch is therefore less a new build than a house move. Anyone who moves packs more than the furniture they like; they also register the new address.
What an existing website contains without anyone having written it down
Before anyone talks about new pages, it is worth looking at what the old website actually does. Five areas are overlooked particularly often.
Addresses. Every URL is a promise to everyone who has saved or linked to it: other websites, industry directories, old newsletters, bookmarks, QR codes on printed material or packaging. If the address changes without a redirect, every one of these routes ends on an error page.
Search visibility. Many visitors do not arrive via the home page but via individual subpages that have been found for specific search queries for years. These are often older guides or product pages that the new concept treats as dispensable. If they are removed or heavily cut, their position disappears with them.
Integrations. Contact forms write to a CRM, newsletter sign-ups go to a mailing service, a shop syncs stock levels with the inventory system, a booking tool creates appointments. Connections like these were often set up once and then forgotten, until they are missing after the relaunch.
Measurement. Analytics code, conversion goals and the configuration of the consent banner determine whether the success of the new site can be judged at all. If they are missing in the first weeks, so is the very before-and-after comparison that matters most.
Legal information. The legal notice and privacy policy have to match the services that are actually embedded. A relaunch often changes those services, yet the texts are carried over unchanged.
What the inventory before a website relaunch looks like
The inventory is not a formality; it is the basis for almost every later decision. It starts with a complete list of all URLs. A crawler, the existing sitemap and exports from Search Console are enough for this; for shops, category and product addresses from the shop system are added.
Each URL needs three pieces of information: how many visits it had over the past twelve months, whether external sites link to it, and which search queries it is found for. From this, every address receives one of four decisions: keep it, merge it with another page, redirect it to a suitable new page, or remove it deliberately. Removing is allowed, but it should be a decision rather than an accident.
In parallel, every form is actually submitted once and traced to where the data ends up. That sounds trivial, but it regularly uncovers connections nobody on the team still knew about. The same applies to tracking code and embedded third-party services.
The result is a list that does not look pretty, but describes the scope of the project more honestly than any moodboard. For a company website with a few dozen pages, the effort is manageable. For shops with thousands of product URLs it is considerable, and it belongs in the schedule and budget accordingly.
Why redirects are not a side issue
Redirects are the part of a relaunch most often handled too late or too crudely. A permanent redirect with status code 301 tells browsers and search engines that a page has moved and where to. Used correctly, it preserves a large share of existing visibility.
Everything to the home page. The most common mistake is redirecting every old address to the new home page. Technically that is a redirect; in substance it is not. Someone looking for a specific article lands on a page that has nothing to do with their question, and search engines treat such redirects much like a missing page.
Chains. If the site has been rebuilt before, there are often old redirects pointing to addresses that are now moving themselves. The result is chains across several hops that cost loading time and eventually break. During the relaunch, all rules should be adjusted so that every old address arrives directly at its destination.
Removed too early. Redirects are occasionally deleted after a few months because they are seen as legacy clutter. External links and printed material do not disappear, though. Redirects therefore belong permanently in the configuration. How structure and search visibility relate in general is covered in the article SEO and Web Design.
When a website relaunch is the wrong answer
Not every problem with a website can be solved by a new website. A relaunch is expensive, ties up the team for months and creates exactly the risks described above. In some situations it is simply the wrong measure.
The offer is unclear. If visitors do not understand what a company offers and for whom, the cause is usually positioning and copy, not layout. A new design carries the same lack of clarity in a more modern form.
Reach is missing. A website that hardly anyone visits will not get more visitors through a relaunch. Content, search engine work or campaigns help more here than a new look.
The problem is narrowly defined. A slow checkout, an unusable form or poor display on mobile devices can often be fixed in a targeted way. A complete rebuild for a partial problem is rarely proportionate.
Nobody maintains the new site. Many websites age not because of the technology but because nobody touches them after launch. Without clear responsibility and time for upkeep, the new site will be in the same state as the old one within a few years.
A relaunch does make sense when the technical foundation no longer receives security updates, when content can only be changed with considerable effort, or when the brand has changed fundamentally. In those cases, carrying on with the old system is riskier in the long run than making the switch.
How changeover day should be run
The changeover itself is short; the preparation is not. A few days beforehand, content on the old site is frozen so that no further changes get lost. Until then the new site runs on a protected staging environment, where redirects, forms and tracking have already been checked in full at least once.
A classic mistake concerns precisely this staging environment: it is blocked for search engines, and the block is carried over at go-live. The new site is then online but invisible to search engines, and this often goes unnoticed for weeks. Checking robots.txt and meta tags therefore belongs on the first-day checklist, together with submitting the new sitemap.
A good time is a morning early in the week when everyone involved is available, not the Friday afternoon before a long weekend. In case something fundamentally does not work, it should be agreed in advance how the old site will be restored and who makes that call.
In the weeks that follow, monitoring is closer than usual: error pages in the server logs and in Search Console, incoming enquiries compared with the previous period, the positions of the most important pages. Fluctuations in search are normal after a changeover. A lasting drop for individual pages, by contrast, almost always points to a missing or incorrect redirect.
After the website relaunch, operations begin
A relaunch is not an end point. The new site has to be maintained, updated and developed further, and the inventory from the start is a good basis for that: it shows which pages carry weight, which systems are connected and who is responsible for what. Keeping that list up to date may make the next major relaunch unnecessary, replacing it with gradual improvement.
How we plan, build and look after websites and commerce systems is described under Services.