The pattern is familiar. New site launches, everyone's pleased, and eight weeks later organic traffic is down forty percent and nobody can say why.
It's avoidable. It's also avoidable in a specific order, and most of the work happens before design starts.
Before you change anything
Crawl the site. Every URL. Export it and keep the file somewhere it won't get lost. You cannot reconstruct this later.
Benchmark. Current rankings, traffic by page, conversions by page. If you don't know what you had, you can't tell what you lost, and you'll spend three months arguing about whether the drop is real.
Find out which pages actually matter. Pull your top pages by organic traffic and by backlinks. These are the pages that need protecting. Frequently they're not the pages anyone in the room cares about. A blog post from 2019 quietly earning a third of your organic traffic is a common discovery, and it's usually not on anyone's redesign plan.
Export your backlinks. Know which URLs other sites point at. Those URLs are assets. Breaking them throws away years of accumulated authority that you can't buy back.
During design
This is where SEO gets decided, which is why bringing it in at the end doesn't work.
Keep URLs where you can. The safest redirect is the one you don't need. There's rarely a real reason to change a URL structure that works. "The new URLs are cleaner" is not worth the risk.
Protect your ranking pages. If a page ranks, be careful rewriting it. Search engines rank it for what it currently says. Change the title, the headings, and half the body, and you've made a different page. Refresh ranking pages deliberately and separately, not as part of a bulk content sweep.
Watch the navigation. Simplifying the nav feels like an improvement and it silently removes internal links. A page that used to be linked from every page and is now three clicks deep will drop. If you're pruning navigation, decide consciously which pages lose their links.
Don't bury content in interactions. Content that only appears after a scroll trigger or inside an accordion still needs to be in the page source.
The redirect map
Every old URL gets a destination.
- 301, not 302. Temporary redirects don't pass authority.
- Direct, not chained. Old to new, one hop.
- To the closest equivalent, not the homepage. A blanket homepage redirect tells search engines the page is gone and wastes whatever it had.
- Test all of them. On staging. Not a sample.
Launch day
Check robots.txt. The single most expensive five-second mistake in this business is launching with the staging block still in place. It takes your entire site out of the index, and it can take weeks to fully recover.
Also: submit the new sitemap, verify Search Console, confirm analytics is firing, and spot-check redirects on the live site.
After launch
Watch for eight weeks minimum.
- Rankings on your benchmarked terms
- Crawl errors and 404s in Search Console
- Traffic by page, compared to your benchmark
- Any 404 that has inbound links
A dip is normal. Recovery in four to eight weeks is normal. Still down after two months is not settling, it's broken, and everything above gives you the list of places to look.
The uncomfortable part
Most traffic loss after a redesign isn't technical. It's editorial.
Somebody decided the old content was embarrassing and rewrote it. The old content was ranking. The new content is better writing and a different page.
If a page earns real traffic, the redesign's job is to not break it. Improving it is a separate project with its own risk.
Pragmatica has been redesigning websites for Canadian non-profits, charities, and associations since 2004. If you're planning one, get in touch.
