There is a specific kind of silence that follows a bad redesign. The site looks current, everyone is pleased, and then the enquiries stop. Two months later someone checks the analytics and finds traffic fell off a cliff on launch day.
This is common, it is preventable, and it almost never has anything to do with the design. It happens because a redesign is treated as a visual project when it is also a migration.
Why traffic actually drops
Search engines have spent years building a map of your site. Every page they know about has an address, and some of those pages have earned the right to appear in results. A redesign usually changes three things at once:
The addresses change. /services/web-design becomes /what-we-do/websites. To a search engine the old page did not move, it vanished. Everything it had earned goes with it unless you say where it went.
The content thins out. A page that ranked because it answered a question in detail gets replaced by a beautiful section with eleven words on it. It no longer answers the question, so it no longer ranks for it.
The technical signals reset. New site, new structure, and often a new sitemap that omits half of what used to be indexed. Sometimes the staging site's instruction to stay out of search gets carried into production, which hides the whole site from Google on day one. That one is more common than anyone admits.
The checklist that prevents it
Whoever builds your site, this is what should happen. If a studio cannot describe this process, that is your answer.
Before launch
- Export every address the old site has. Not just the pages in the menu. Pull the full list from the sitemap, the analytics, and Search Console, because the pages that earn traffic are often not the ones anyone remembers.
- Find out which pages actually matter. Sort by visits and by search impressions. A handful of pages usually carry most of the value. Those are the pages you protect.
- Keep the content that earns. If a page ranks because it answers something thoroughly, the new version needs to answer it at least as thoroughly. Redesign the presentation, not the substance.
- Write the redirect map. Every old address gets a permanent redirect to its closest equivalent. Not all to the homepage, which tells the search engine the content is gone. To the actual matching page.
- Check the staging site is blocked and the live one is not. Then check again on launch day.
At launch
- Redirects live and tested, including the awkward ones with query strings or trailing slashes
- Sitemap regenerated and submitted
- Analytics and Search Console still connected, so you can see what happens
- Internal links updated to point at the new addresses rather than bouncing through redirects
After launch
- Watch for two weeks, not two days. Some fluctuation is normal while the index updates. A sustained drop is not.
- Check the crawl errors. Search Console will tell you exactly which addresses it can no longer find. That list is your fix list.
- Compare like for like. Traffic to the same pages, before and after, rather than a total that a seasonal dip can disguise.
When a redesign is not the answer
Sometimes the site is fine and something else is wrong. Before rebuilding, it is worth knowing which of these you actually have:
- It looks dated but converts. Consider whether the money is better spent on the pages that carry the buying decision rather than the whole site.
- It is slow. Speed is often fixable without a rebuild. Oversized images, bloated plugins and render-blocking scripts are usually the culprits, and they are cheaper to fix than to redesign around.
- Nobody can find it. That is a content and structure problem. A new coat of paint on the same thin pages changes nothing.
- It breaks on phones. Sometimes genuinely fixable, sometimes a symptom of a build that has run out of road. Worth an honest look before committing.
A redesign is the right answer when the site actively works against the business: when the structure no longer matches what you sell, when nobody on your team can update it, or when the underlying build makes every change expensive.
The uncomfortable part
The most common cause of a lost-traffic redesign is that nobody was responsible for the migration. The designer designed, the developer built, and the redirect map was somebody else's problem, which meant it was nobody's.
So ask the question before you sign, alongside what the whole thing should cost: who is handling the redirects, and how will we know it worked? A studio that has done this before will answer immediately and specifically. That answer is worth more than a portfolio.
If you want a second opinion on whether your site needs rebuilding at all, send it over. We will look at it properly and tell you honestly, including when the answer is that it does not.