The worst outcome of a redesign is not an ugly website. It is a beautiful one that produces fewer enquiries than the site it replaced, and a team with no idea which change caused it. That happens when a redesign starts with a look rather than with evidence.
These seven checks take a few days and repeatedly save projects from that outcome.
1. Know which pages actually earn
Open your analytics and list the pages that bring in traffic, and the pages that produce enquiries. They are often not the same pages, and they are rarely the ones the team argues about. Anything on that list has earned the right to survive the redesign, or at least to be replaced deliberately rather than by accident.
2. Check the tracking before you trust it
Before drawing conclusions, confirm that analytics is recording what everyone assumes. Duplicate tags, missing goals, form submissions never fired, bot traffic counted as visits — we find at least one of these in most audits. Rebuilding a site based on broken data is expensive.
3. Write down why you are redesigning
“It looks dated” is a symptom, not a goal. Specific goals sound like: more enquiries from the services pages, a site the marketing team can update without a developer, faster pages on mobile, a checkout that loses fewer people. Write them down before the first design, because they are what you will judge the result against.
If you cannot measure it, you cannot claim it
Take a baseline before anything changes: traffic, enquiries, conversion rate, Core Web Vitals, and the rankings of your top twenty pages. After launch, this baseline is the only thing that can tell you whether the project worked.
4. Map every URL
Export a full list of your current URLs. For each one decide: does it stay the same, move to a new address, or disappear? Every move needs a redirect. This single spreadsheet is what protects the search traffic you have spent years accumulating, and it is the step most often skipped.
5. Audit the content honestly
A redesign is the cheapest opportunity you will ever get to delete pages nobody reads. Go through the list: keep, rewrite, merge or remove. Moving thin, outdated content onto a shiny new design just makes it more visible.
6. Find out where visitors give up
Session recordings and funnel data show the hesitation your analytics summary hides: the form field people abandon, the price they hunt for and never find, the section everyone scrolls past. Fixing those specific moments usually matters more than the visual refresh.
7. Decide who owns the site afterwards
Who updates content? Who applies security updates? Who is called when something breaks? If the honest answer is “nobody”, the new site will look exactly like the old one in three years. Decide this during the project, not after the invoice is paid.
Who should be in the room
Redesigns stall when the wrong people approve them. You want the person who owns the commercial outcome, the person who will update the site afterwards, and someone who talks to customers daily — usually sales or support. That third person knows which questions your pages fail to answer, and they are almost never asked.
You do not want design decisions made by committee. Agree upfront who gives feedback and who makes the final call, and the project will move twice as fast.
What a realistic timeline looks like
For a typical business website, four to eight weeks from discovery to launch is normal. Larger sites, platform changes and content writing push that out. The parts that slip are rarely technical: they are waiting for content, waiting for feedback, and waiting for a decision nobody was assigned.
Ask for a plan with dates and named owners, including your own. A timeline that only lists the agency’s tasks is a timeline that will move.
Mistakes we see most often
- Redesigning because a competitor did, rather than because of a measured problem
- Losing search traffic by skipping the URL map
- Designing with placeholder text and discovering the real content does not fit
- Removing pages that quietly brought in enquiries
- Launching on a Friday afternoon
- Treating launch as the end of the project rather than the middle
One question before you start
If the redesign works perfectly, what number changes? Enquiries, orders, cost per lead, support tickets, time to publish a page — any of them is a fine answer. “It looks better” is not, because in six months nobody will be able to tell you whether the money was worth spending.
After launch: the first month matters
- Verify every redirect actually resolves
- Submit the new sitemap and watch for crawl errors
- Check that analytics, goals and conversion tracking still fire
- Compare rankings for your top pages weekly, not once
- Watch real-device speed, not just a lab score
- Collect feedback from the people who use the site daily
A redesign is not finished at launch. The first few weeks are when you find the problems that no amount of staging review would have caught, and when a quick fix is still cheap.
