Most website redesigns go wrong before a single screen gets designed. Someone opens Figma with a rough sense of “make it look better” and no answer for what happens to the two hundred URLs already ranking, who has final say on the homepage copy, or what “done” actually means. The design work is rarely the hard part. The decisions nobody made before it started are.
Here is the checklist I actually run through with clients before opening a design file, in the order it needs to happen, with the parts people skip and what skipping them costs.
1. Decide why you are redesigning at all
“It looks dated” is a reason, but it is not a brief. Before anything else, write down the actual business problem: conversion is flat, the CMS cannot handle the content team’s workload, mobile traffic bounces at twice the desktop rate, the brand outgrew the site. This sentence becomes the test every design decision gets measured against later, and it is the difference between a redesign that moves a number and one that just moves pixels.
If the honest answer is “the founder is bored of looking at it,” say that out loud too. It is a legitimate reason, but it changes the scope: you are buying a fresh coat, not fixing a structural problem, and the budget and timeline should reflect that.
2. Audit the content that already exists
Before anyone designs a new page, someone needs to know what is on the old one and whether it earned its place. Pull every URL, note traffic and rankings if you have analytics, and sort pages into three buckets: keep, merge, retire. This step alone usually cuts a site’s page count by a third, which means less to design, less to migrate, and a cleaner sitemap for whoever inherits it next.
Skipping the audit is the single most common cause of a redesign running long, because content decisions that should happen in week one instead happen mid-build, one page at a time, while the designer waits.
3. Lock the URL and redirect plan before design starts
This is the step that protects search rankings, and it is the one non-designers forget exists. Every URL that currently ranks needs a decided destination on the new site, even if that destination is “this page no longer exists, redirect to the parent category.” Write it as a spreadsheet: old URL, new URL, redirect type. Hand it to whoever builds the site before launch, not after someone notices traffic dropped.
| Old URL status | What to do |
|---|---|
| Ranking, keeping the content | Map to its new URL, 301 redirect, keep the H1 close to the old one |
| Ranking, content being merged | 301 redirect to the page it merged into, not the homepage |
| Ranking, content being cut | Find the closest relevant new page and redirect there rather than losing the equity |
| Not ranking, low value | Safe to drop, no redirect needed |
| New page, no old equivalent | No redirect needed, just make sure it is in the new sitemap |
A blanket redirect of everything to the homepage is the fastest way to lose months of rankings overnight. I have inherited two sites where that was the previous developer’s shortcut, and both took over a quarter to recover.
4. Set up analytics before, not after, launch
You cannot measure whether a redesign worked if you did not capture a clean baseline first. Confirm your current analytics setup is actually tracking the events that matter, form submissions, calls to action, scroll depth on key pages, and export at least three months of that data before launch. Then make sure the same events are wired into the new site before it goes live, not added as an afterthought once someone asks why the dashboard is empty.
5. Decide who signs off, and on what
Every redesign that drags has the same root cause: nobody agreed in advance who approves what, so every round becomes a fresh negotiation. Before design starts, name one person per decision type. One person approves visual direction. One person approves copy. One person has final say when the two disagree. It does not need to be the same person for each, but it needs to be decided, in writing, before the first mockup, not during round four of “let’s get one more opinion.”
I put a fuller version of this into a working brief template, and the questions that surface the scope people usually miss are worth asking regardless of who is designing the site: I cover them in questions to ask when hiring a UX designer.
6. Confirm the platform before the design does
Staying on the same CMS or moving to a new one changes what the design can assume. A redesign that assumes flexible content blocks will not survive a platform that only offers rigid templates, and the other way around. Settle this before design starts, not after the first Figma file is built around assumptions the platform cannot support. If you are weighing WordPress against a page builder platform like Webflow, I break down the actual tradeoffs in Webflow vs WordPress for business websites.
7. Budget for content, not just design
New page structures usually need new copy, not the old copy reflowed into a new layout. Decide now whether you are writing it, your team is, or the designer is, because “the designer will handle it” quietly becomes hours of unplanned work that were never priced. This is the same gap I flag when people ask what a redesign costs: the design fee and the total project cost are rarely the same number, for the reasons I lay out in what a custom website costs.
8. Plan the launch, not just the build
Decide the cutover date, who tests the new site before it goes live, and what the rollback plan is if something breaks. A staging environment that mirrors production, a pre-launch checklist covering forms, redirects and analytics, and a named person watching the site for the first 48 hours after launch turn a stressful night into a boring one. Boring is the goal.
Frequently asked questions
How long should a website redesign take?
For a business site of ten to twenty pages, six to ten weeks from a decided brief to launch is realistic, longer if the content audit turns up more restructuring than expected. The variable that moves this most is not design time, it is how fast decisions and content get delivered on your end.
Will a redesign hurt my SEO rankings?
Only if the URL and redirect plan is skipped or done carelessly. Done properly, with every ranking URL mapped to a 301 redirect and content preserved or genuinely improved, most sites see a short dip and recover within four to eight weeks, often landing higher than before.
Should I redesign everything at once or in phases?
Phasing works well for large sites: redesign the highest-traffic templates first, homepage and top landing pages, then work outward. For a smaller business site, one coordinated launch is usually simpler and avoids the site looking half-renovated for months.
Do I need new content for every page, or can I reuse what I have?
Reuse whatever is already working, particularly pages that rank well and convert. Rewrite where the content audit shows thin or outdated copy, or where the new structure genuinely needs a different shape of content than the old page had.
What is the biggest mistake people make in a redesign?
Treating it as a visual refresh instead of answering why the old site was not working. A new coat of paint on the same structural problems just means the same complaints in a nicer font, six months later.
If you are about to start one
I am a UI/UX and web designer based in the Philippines, working with clients in the US, Australia and Asia on exactly this kind of project. If you want a second opinion on your redesign plan before you commit to a timeline, have a look at how I work and send me where things stand.
Planning a website or product design project?