Article
Jun 10, 2026
A Conversion-Focused Website Redesign, Step by Step
The operating sequence we run before anyone touches a visual. Baseline math, message tests, evidence-wired wireframes, a speed budget as a design rule
TL;DR
A conversion focused website redesign starts with analytics math, not Figma boards or competitor screenshots.
Lock the message before the layout. Wireframes should follow CRO evidence, not inspiration sites.
Treat page speed as a design constraint with a hard budget, not a developer's problem at the end.
Launch with instrumentation already wired: events, funnels, session recordings, a baseline you can compare against.
Plan the first 90 days of tests before launch day, not after the agency invoices clear.
If you're searching for a conversion focused website redesign process that does more than swap fonts, here's the short answer: spend the first 3-4 weeks before any visuals on baselining current conversions, testing the message, and writing wireframes from evidence. The visual phase is the easy part. The work that decides whether you redesign a website without losing conversions, or quietly bleed leads for two quarters, happens before a single pixel moves.
1. Why most redesigns are repaints that quietly lose leads
Most agency redesigns are a repaint with a new typeface. The header gets bigger, the hero gets a video, the CTA changes color from blue to orange, and three months later the marketing director is staring at a form-fill chart that drifted down 18% and nobody can explain why.
The cause is almost always the same: the redesign was scoped as a visual project, when the thing actually being changed was the conversion path. New navigation hides the proof points buyers used to read. The new hero buries the offer the old hero was carrying. The fold moves. The form moves. The trust signals get demoted to the footer because they ruined the layout. Each decision was defensible in isolation. Together, they broke the page's job.
A cro redesign process treats the page as a system with measurable inputs and one measurable output, then redesigns the system. The visuals are downstream of that. Beautiful is a constraint, not the goal.
2. Step 1: Baseline: analytics, session recordings, current conversion math
Before we sketch anything, we write down what the current site actually does. Not what the team thinks it does, what GA4, the CRM, and 50 session recordings prove it does.
The baseline document we hand the client usually covers six things:
The baseline is the document that lets you, six months from now, prove the redesign worked. If you skip it, every post-launch conversation becomes vibes versus vibes. We won't ship without it. (For a sense of where the budget lands at this stage, see our note on website redesign cost for small business.)
3. Step 2: Message before layout: what the page must say
A layout is just a container for a message. If the message is wrong, a better container makes the loss faster.
We spend roughly a week on message before we touch wireframes. Three artifacts come out of it:
A one-sentence value proposition the founder can say out loud without cringing. It names the buyer, the outcome, and the mechanism.
A problem-to-solution overview, 200-300 words, that earns the rest of the page. The structure is borrowed from Rattleback's service-page research, which argues for that length and for picking few, deep case studies over many shallow ones. We've found the same pattern holds in our client work, particularly on B2B service pages where the buyer is doing diligence, not browsing.
A proof inventory: the actual numbers, named clients, certifications, and quotes the page can stand on. If the inventory is thin, we don't paper over it with stock photography. We tell the client what to go gather before the page can carry weight.
4. Step 3: Wireframe from CRO evidence, not inspiration sites
We don't open Dribbble. We open the baseline document and the message brief, and we wireframe from those.
The sequence on a typical service page, in order of the buyer's question:
Who is this for, and what's the outcome? (Hero, 8-12 words of headline, one supporting line.)
Do you understand my problem? (200-300 word overview, per the Rattleback pattern.)
How do you actually do this? (Mechanism, the steps, the deliverables, the timeline.)
Has this worked before, for someone like me? (Two or three deep case studies with real numbers, not a logo grid.)
What does it cost, roughly, and what happens next? (Pricing range or qualifying questions, then the form.)
5. Step 4: A speed budget as a hard design constraint
Speed is a design decision, not a developer's clean-up task. So we set a budget before anyone exports an asset.
The budget we give designers on most projects: under 1.5 seconds Largest Contentful Paint on a mid-tier mobile device on 4G, page weight under 500KB above the fold, no more than one custom font family with two weights. Hero video gets a hard look. Most heroes don't need it.
The stakes here are not theoretical. Deloitte's 2020 Milliseconds Make Millions study, 37 brands, roughly 30 million sessions, found that a 0.1-second mobile speed improvement lifted lead-generation form submissions by 21.6%. One tenth of one second. That is the size of the lever, and it's the one most redesigns hand back to the carousel plugin.
6. Step 5: Launch with a measurement plan, not a party
Launch day is when most agencies disappear. It's when our work gets interesting.
Before the DNS flips, four things have to be true:
GA4 events match the old taxonomy where possible, so the baseline is comparable. Where events had to change, the mapping doc is written and the client has it.
The CRM is receiving form submissions on the new page in staging, with the right source attribution. We've seen too many launches where forms worked but lost their UTM tags.
Session recording and heatmap tools (Hotjar, Microsoft Clarity, whatever the client uses, see each tool's published pricing page) are live on day one, not added in week three.
A 301 redirect map covers every URL that's changing, with the SEO checks the team owns. If you haven't run one before, our website redesign SEO checklist walks the discipline.
7. What we'd test in the first 90 days after launch
Launch is a hypothesis, not an answer. The first 90 days is where you find out which of your wireframe bets were right.
A reasonable test cadence for a mid-traffic B2B site (10-30k monthly sessions):
Webflow's vendor-published Optimize case study claims Walker & Dunlop saw a 56% lift in form fills from structured post-launch testing. The methodology isn't shown, so treat that as directional, not a benchmark to underwrite a budget on. In our client work, sustained testing in the first quarter typically moves conversion rate by 15-40% off the new baseline. Sometimes more. Always less than the vendor case studies promise.
FAQ
How long does a conversion focused website redesign actually take?
For a 15-30 page B2B site, the sequence above runs 10-14 weeks in our experience: 3-4 weeks on baseline and message, 2-3 weeks on wireframes, 4-5 weeks on visuals and build, then a 1-2 week instrumented launch window. The 90-day test loop runs after that. Anything faster usually skips the baseline, which is the step you can't recover.
How do I redesign a website without losing conversions during the cutover?
Three disciplines do most of the work. A complete 301 redirect map covering every old URL. GA4 events that match the old taxonomy where possible so baselines stay comparable. And form-submission testing in staging with real CRM source attribution. If those are wired before launch, a week-one dip becomes diagnosable instead of mysterious.
Is page speed really worth designing around?
Yes, and the lever is bigger than most teams assume. Deloitte's 2020 Milliseconds Make Millions study, covering 37 brands and roughly 30 million sessions, found a 0.1-second mobile speed improvement lifted lead-gen form submissions by 21.6%. That's why we set a speed budget before the design starts, not after the developer audits the build.
What's wrong with starting from inspiration sites?
Inspiration sites optimize for portfolio screenshots, not for your buyer's questions. They're built for a different audience under different constraints. Starting from CRO evidence, your analytics, your session recordings, your message brief, produces a page that answers the questions your actual buyers ask in the order they ask them. The visuals come later, and they come easier.
When should we plan the post-launch test program?
Before launch, not after. Every wireframe block in our process gets a hypothesis attached, what change we expect, what direction, what rough magnitude. Those hypotheses become the first 90 days of A/B tests. Planning the program before launch means the testing tool is wired on day one and the first test ships in week three, not month three.
Ship the baseline document this week. Watch 30 session recordings on the highest-value entry page, write down what you see, and bring it to whoever is briefing the redesign. That single artifact will change the brief more than any agency pitch. When you want a second set of eyes on it, get in touch.
How should a small team prioritize digital marketing agency for web design and seo?
Start with the workflow that already has a baseline: hours, leads, errors, or budget waste.
What should be measured before investing in digital marketing agency for web design and seo?
Measure cycle time, volume, handoffs, error rate, and the current owner.
When should conversion focused website redesign process stay manual instead of automated?
Keep it manual when judgment, approval, brand nuance, or customer trust is on the line.
How does digital marketing agency web design change the budget for digital marketing agency for web design and seo?
digital marketing agency web design usually adds integration, QA, and monitoring work.
What is the first project to launch from this conversion focused website redesign process playbook?
Launch the narrowest workflow with a visible result.
Related reading
The 5-Tool AI Stack for Small Business, With Real Monthly Costs
AI Search Optimization for Small Business: What AEO Costs (and What to Do Free at DR 0)
AI SEO Services Pricing: What We Charge and Where the Money Actually Goes