Six weeks after the new site launches, someone notices the phone is quieter. Traffic is down by half. The site looks better than it ever has.
This happens often enough that fear of it stops people commissioning rebuilds they need. So it’s worth naming what actually causes it, because it’s rarely mysterious and it’s almost entirely preventable.
It’s usually not the design. It’s the addresses.
What actually breaks
Every page on your current site has a URL, and Google has spent years associating each one with what it’s about and how much to trust it.
Rebuild the site and those URLs often change. A page that lived at /services/drain-cleaning becomes /drain-cleaning-services. Google arrives at the old address, finds nothing, and the accumulated history of that page has nowhere to go.
Do that across forty pages and you haven’t redesigned your site. You’ve launched a new one and quietly deleted the old.
The fix is a redirect: a rule saying the old address now points at the new one. It passes most of the history across. It takes minutes per page and it is the single most skipped step in the industry.
It also fixes a problem you cannot see from inside your own business: every link anyone else ever made to you. Directory listings, a supplier’s partner page, an old chamber of commerce entry, a customer’s blog post. Those links point at addresses you are about to retire, and without redirects every one of them lands on an error page. You did not build those links and you cannot rebuild them.
The other three causes, in order of frequency
The staging site was never unblocked. Sites are built with a flag telling search engines to ignore them. It is supposed to be removed at launch. It frequently isn’t, and the site is invisible until someone notices.
This is the most embarrassing and easiest to check, and it should be the first thing you look at if traffic goes to nearly zero rather than just down.
Content got shorter. Redesigns tend to trim. A page that had eight paragraphs answering real questions becomes three lines and a nice photo. It looks cleaner and it ranks worse, because the thing that was ranking was the answer, not the layout.
Minimal design and minimal content are different decisions, and conflating them is a common error. The case in why minimalist, conversion-focused design wins is about removing distraction, not removing substance.
The site got slower. New builds sometimes ship with oversized images and more scripts than the old one. Speed affects both ranking and whether anyone waits around.
Want a second opinion on this?
Book a free consultationWhat to inventory before launch
Four things, and they take an afternoon.
Every current URL. A full list. Your sitemap gives you most of it, and analytics gives you the ones that get traffic. This list is the raw material for redirects.
Your top twenty pages by traffic. These are the ones where a mistake actually costs money. If time is short, protect these first.
Current rankings for your main terms. Write them down. Without a before, you cannot tell whether an after is bad.
Every page’s title and description. They get regenerated during a rebuild, often into something generic.
Hand that inventory to whoever is building. If they don’t ask for it, ask them how they plan to handle redirects, and treat a vague answer as the finding it is.
What to check in the first week after
Whether the site is indexable at all. Check for a leftover block first.
Whether the old URLs redirect. Take twenty from your list, type them in, and confirm each one lands somewhere sensible rather than on an error page or the homepage. Redirecting everything to the homepage is technically a redirect and practically useless.
Whether the pages you cared about still say what they said. Compare against your inventory.
Then watch for four weeks before concluding anything. Some movement is normal while Google re-crawls, and panicking in week one usually causes worse decisions than waiting.
One caveat on that patience. Four weeks is the right window for ranking movement, not for outright errors. If the site is blocked from indexing, or old URLs are returning errors, fix those the day you find them. The waiting advice applies to interpreting numbers, not to leaving something broken in place while you wait to see how broken it gets.
If traffic was already thin before the rebuild, the rebuild isn’t your problem, and the question is whether you needed one at all: when a site needs a rebuild and when it needs three fixes.
If you’re about to commission a rebuild, ask how redirects will be handled before you sign anything. The answer tells you a lot.
Frequently asked questions
How long does recovery take if redirects were missed?
Adding them late usually recovers most of it, though it can take weeks to re-crawl. The longer the gap, the more ground to make up.
Should I keep my old URLs instead of changing them?
Often yes. If the existing structure is sensible, keeping it removes the whole risk. Change URLs when there’s a reason, not as a side effect of a new theme.
Does changing platform make this worse?
It makes URL changes more likely, because different systems structure addresses differently. Same risk, higher probability.
Can I do this myself?
The inventory, yes, and it’s worth doing yourself even if someone else handles the redirects. It’s your list of what matters.
