Website Redesigns That Don't Tank Conversion

Most B2B website redesigns produce a better-looking site that converts worse. The new design is cleaner, more modern, visually coherent. The new information architecture is more rational. The new brand positioning is more polished. And the conversion rate drops, the organic traffic drops, and the pipeline follows. Six months after launch, the team is quietly re-adding elements they'd removed during redesign because the removed elements were doing work that wasn't obvious until they were gone.

This pattern is so common it's almost a default. It happens because redesign projects are set up to optimise for the visible output (the new site) rather than for the commercial outcome (conversion and pipeline). When a project's success criteria is "launch a better-looking site by Q3," the site launches and the conversion damage emerges post-launch as an unrelated problem. When the success criteria is "launch a better-looking site that maintains or improves conversion," the discipline that protects conversion gets built in from the start.

This article is about what that discipline looks like, the specific failure modes that tank conversion during redesigns, and the sequence that actually produces a redesigned site that performs as well or better than the one it replaced.

Why Most Redesigns Hurt Conversion

Four specific patterns that reliably damage conversion during redesigns.

Removing elements that were doing invisible work. The old site had a testimonial slider, a feature grid, a pricing calculator, a demo video. The redesign removes them because they clutter the visual composition. Each individual removal seems fine. Collectively, they strip out the conversion scaffolding that was moving visitors toward action. The new site looks cleaner and converts less because the cleanness came at the cost of what was actually producing pipeline.

Optimising for aesthetic coherence over conversion-specific patterns. Good design wants visual coherence, consistent spacing, unified typography, balanced layouts. Conversion often requires breaks from this coherence, a specific callout that disrupts the pattern, a CTA with unusual prominence, a form that's deliberately visually heavy. Redesigns led by design teams without conversion input tend to smooth out these patterns, which smooths out the conversion performance.

Breaking URL structure and inbound SEO. The new site has a better information architecture, which means new URLs. Old URLs 301 to new ones (or should; often they don't). Inbound links from years of SEO work point at URLs that now redirect or break. Rankings that took 18 months to build drop within weeks because the URLs they pointed to no longer exist in the shape Google's index expected.

Replacing proven copy with positioning-polished copy. The homepage copy that the old site had was specific, testable, and probably optimised over time based on performance feedback. The new copy is a result of the positioning exercise the redesign triggered, more coherent with the brand, more polished in tone, and often more abstract. Abstract copy converts worse than specific copy, but abstract copy feels like a positioning upgrade rather than a conversion downgrade.

What Protects Conversion Through Redesign

Five disciplines that, combined, produce redesigns that maintain or improve conversion rather than degrading it.

1. Baseline the current site quantitatively before touching anything

Before the redesign starts, capture the conversion performance of the current site. Not just the headline number, segment-level performance. Conversion rate by traffic source. By page. By device. By visitor intent. The goal is a detailed baseline that the new site can be measured against, page by page.

Without this baseline, "the new site is working" becomes a matter of opinion. With the baseline, the team can tell which pages improved, which stayed flat, and which degraded, which is the information that drives post-launch optimisation.

2. Preserve the high-converting pages exactly

Some pages on the current site are converting well. Sometimes for reasons you can identify, sometimes for reasons that aren't obvious. The redesign should preserve these pages' core structure, the content, the flow, the specific elements that are doing the converting work, even if the visual design changes.

The instinct is to redesign everything because consistency. The better instinct is to redesign everything except the 3-5 pages that are producing most of the conversion. Those pages get a visual refresh that's explicitly subordinated to not breaking what's working.

Which pages are these? Usually the homepage, the highest-trafficked product or service pages, the highest-converting landing pages. Pulling analytics to identify the specific top-converting pages and treating them as preservation targets rather than redesign canvases protects the bulk of conversion through the transition.

3. Run URL mapping and redirect validation before launch

Every URL on the current site maps to a URL on the new site. Either the URL stays the same, or there's a 301 redirect from old to new. No URL on the current site, especially pages with inbound links or search traffic, should 404 on launch.

The URL mapping exercise is boring and essential. A spreadsheet with every current URL, its traffic and inbound link count, and its new destination. Pre-launch validation that every redirect actually fires correctly. Post-launch monitoring for 404s in the first month to catch anything missed.

Teams that skip this exercise lose SEO traffic post-launch and struggle to diagnose why. Teams that do it thoroughly preserve the organic inbound the old site had earned.

4. A/B test the most consequential page changes where possible

The homepage, the pricing page, the main contact form. If the new version of these pages is meaningfully different from the old version, running a split test before full rollout surfaces performance differences before the conversion damage is locked in.

Not every page can be tested, content pages rarely justify the setup. But the 5-8 most-visited and highest-converting pages usually can be. Running them as tests means the new version has to demonstrate it performs as well as the old version before fully replacing it.

This discipline changes the launch pattern from "redesign launches as a complete replacement on a specific date" to "redesign launches in waves as each high-value page's new version is validated." The waves take longer. The outcomes are reliably better.

5. Keep the old version accessible for reference for 6 months

Archive the full current site before launch. Keep it accessible internally (not publicly) for 6 months post-launch. When questions arise about why the old site had a specific element, "didn't there used to be a pricing calculator on this page?", the archive provides the answer.

This archive matters because the invisible-work-of-removed-elements only becomes visible after it's missing. Having the old version accessible means the team can identify what was removed and, if necessary, restore specific elements that were doing work.

The Sequence That Works

A rough sequence for a B2B website redesign that preserves conversion. Assumes a 4-6 month project.

Month 1: Baseline and analysis. Capture current performance. Identify preservation targets. Build the URL mapping. Name the conversion disciplines the redesign will adhere to.

Month 2: Strategy and positioning work. If the redesign is triggered by a positioning shift, do the positioning work properly here rather than embedding it into the design process. A clear positioning document before wireframes prevents positioning changes from confusing the design process.

Month 3: Wireframes and structure. Design the new information architecture. Wireframe each page type. Explicitly identify which pages will be preserved structurally and which will be redesigned. Get sign-off on the wireframes before visual design starts.

Month 4: Visual design and build. Apply the new visual language. Build the pages. A/B test setup for the critical pages. URL redirect configuration.

Month 5: Pre-launch validation. URL redirect testing. Performance testing of the new pages. Content review. Final SEO checks. Stakeholder review.

Month 6: Phased launch. Launch non-critical pages first. Run A/B tests on the critical pages. Roll out the full site as tests validate. Monitor conversion and organic traffic daily for the first month post-launch.

Shorter timelines compress each phase, which is where the errors happen. Faster redesigns damage conversion more reliably than slower ones.

What Redesigns Often Should Not Include

Three things teams often bundle into redesign projects that would be better done separately.

Positioning rework. Redesigns often trigger positioning work because the new site needs copy. Doing positioning work inside a redesign project produces rushed positioning that gets baked into the new site. Better: finish positioning work before the redesign starts.

Major functional changes. New tools, new integrations, new site functionality. Each adds scope and risk. Redesigns that include significant new functionality usually ship late or ship with bugs that tank conversion for the first month.

Complete brand overhauls. New logo, new colour palette, new typography, new voice. Individually each is manageable. Combined, they produce a site so different from the prior version that returning visitors don't recognise it and new visitors form first impressions of an unfamiliar brand.

Each of these is legitimate work. None should be stacked on top of a redesign project. The redesign becomes the coordination disaster when too much other change is routed through it.

Where This Sits In The NBT Growth Model

Website redesign is a Navigation-stage execution activity, but the risk profile makes it worth treating with Checkup and Formulation disciplines first. The diagnostic work (what's currently converting, what's not) and the positioning work (what the site should actually say) belong upstream of the redesign rather than embedded in it.

The connection to pillar-and-cluster SEO is direct. Redesigns often disturb the SEO foundation the business has built. Teams that understand their organic traffic patterns and protect them through redesign preserve the compounding asset; teams that don't often spend two quarters rebuilding what they inadvertently broke.

If you're planning a website redesign and want to make sure it doesn't tank your conversion or SEO, or you've recently launched a redesign and the numbers have dropped, get in touch for a strategy call. We'll walk through the specific disciplines that apply to your situation and help identify what to protect.