Most redesigns look better and perform worse. The new site launches, everyone congratulates each other, and three months later leads are down 30% with nobody quite sure why.
Here's what usually went wrong.
1. Redesigning Because the Site Looks Dated
"We're bored of it" is not a business case. Your customers see your homepage once or twice; you see it every day. Familiarity breeds contempt in the marketing team long before it does in the market.
Before redesigning, find the actual problem. Is traffic down? Conversion rate down? Are support tickets full of people who can't find pricing? Redesign to fix a measured problem, and you'll know whether it worked.
2. Not Benchmarking Before You Start
You cannot prove a redesign succeeded if you never recorded where you started.
Capture, at minimum: sessions and sources by month, conversion rate per key page, top 50 landing pages by traffic and conversions, current keyword rankings, and average page load time. Screenshot everything. This takes an afternoon and it's the only thing that will let you diagnose problems later.
3. Losing Your URL Structure
This is the big one. Changing URLs without 301 redirects is the fastest way to delete years of SEO equity.
Every old URL that changes needs a permanent redirect to its closest new equivalent. Not to the homepage — to the relevant page. A homepage redirect for hundreds of URLs gets treated as a soft 404 and passes almost nothing.
Crawl the old site before launch, export every URL, and build the redirect map as a deliverable. Then verify it after launch, not before.
4. Designing for the Homepage
Teams spend 60% of the design budget on a page that most visitors never see. Organic search, ads, and email all drop people onto interior pages.
Look at your landing page report. Design the top five entry points with the same care as the homepage.
5. Removing Content That Was Ranking
That old blog post with the ugly formatting may be responsible for a third of your organic traffic. Auditing content by how it looks rather than how it performs is how sites lose visibility overnight.
Pull a performance report first. Anything driving traffic, links, or conversions gets migrated and improved — not deleted.
6. Hiding the Call to Action Below the Fold
Modern hero sections have grown enormous. Giant image, giant headline, lots of whitespace — and the primary action pushed off-screen on a laptop.
Whitespace is a design tool, not a virtue. If a visitor has to scroll to discover what you want them to do, you've added friction for no benefit.
7. Replacing Clear Copy With Clever Copy
"Empowering tomorrow's possibilities" tells a visitor nothing. "Commercial plumbing repairs across Manchester, 24/7" tells them everything.
Your headline has about three seconds to answer: what is this, is it for me, and what do I do next. Cleverness costs you seconds you don't have.
8. Ignoring Page Speed Until After Launch
Full-width video backgrounds, four webfont families, an animation library, and unoptimized hero images — each defensible alone, collectively fatal.
Set a performance budget at the start: target Largest Contentful Paint under 2.5 seconds on a mid-range phone over 4G. Test at the design stage. It's far cheaper to reject a heavy design concept than to strip it back after it's built.
9. Skipping the Mobile Design Entirely
Designing on desktop and letting the framework "handle mobile" produces sites that technically work and practically don't — tap targets too small, forms that trigger the wrong keyboard, sticky headers eating half the viewport, tables that scroll sideways.
With most traffic on mobile, the phone layout deserves deliberate design attention, not automatic reflow.
10. Adding Friction to Forms
Every additional field costs you conversions. Ask for what you need to make the next contact, nothing more.
Common offenders: requiring a phone number for an email newsletter, mandatory company size fields, CAPTCHAs that fail on mobile, and forms that clear all entered data when validation fails.
11. Launching Without Checking Tracking
Analytics tags get lost in migrations constantly. So do conversion events, ecommerce tracking, ad platform pixels, and CRM form connections.
The result is a fortnight of blind operation followed by the discovery that you have no data for the most important period of the project. Build a tracking verification checklist and run it within an hour of going live.
12. Treating Launch as the Finish Line
The redesign is a hypothesis. Launch is when you start testing it.
Plan for a six-week post-launch phase: monitor rankings weekly for redirect issues, watch Search Console for crawl errors, compare conversion rates against your benchmarks, run session recordings to see where people hesitate, and fix what the data exposes.
A Minimum Launch Checklist
Before you go live:
- Full redirect map built and tested
- Analytics and conversion events verified on the new site
- XML sitemap regenerated and submitted
- robots.txt checked — staging blocks removed
- Meta titles and descriptions migrated for ranking pages
- Forms tested end to end, including the notification email
- Core Web Vitals measured on real devices
- Accessibility audit passed
- 404 page in place and useful
- Full backup of the old site retained
Expect a Dip
Even a well-executed migration usually sees a two-to-six-week ranking wobble while search engines reprocess the site. That's normal.
What's not normal is a decline that persists past eight weeks. That's a redirect problem, a content problem, or an indexing problem — and it's fixable, but only if you catch it early. Which brings us back to mistake number two.
Planning a redesign? We run a pre-launch technical audit that catches most of these before they cost you traffic. Get in touch before you build, not after.