The website redesign checklist nobody sends you until it is too late

Most redesign projects go wrong in the same two places: the content that was supposed to arrive in week two, and the redirects that nobody thought about until launch week. Both are avoidable.

Here is what to settle before the work starts.

Before you brief anyone

1. Write down why you are redesigning. "It looks dated" is a reason; "enquiries have fallen and we do not know why" is a different project with a different answer.
2. Decide what success looks like, in a number you already measure. More enquiries, lower bounce on the pricing page, fewer support calls asking where something is.
3. Export twelve months of analytics. Which pages get traffic, which convert, which nobody has visited since 2023.
4. Export sixteen months of Search Console data. Which pages earn impressions and clicks, and for what.
5. Pull your backlinks. Any page with an external link pointing at it needs to survive the rebuild, even if it gets no traffic.
6. Agree who the single decision-maker is. Design by committee is the most reliable way to double a timeline.

Content

1. Audit every existing page: keep, rewrite, merge or delete. Do this before design, because it changes the page count.
2. Decide who writes the new copy, and put dates against it. This is the item that slips.
3. Check image rights. Stock licences do not always transfer to a new site, and "we found it on Google" is a liability.
4. Gather what you already have: real photographs, product shots, team pictures. Real imagery beats stock every time and costs nothing extra.
5. List what you need that does not exist yet. Photography and video have lead times that ruin launch dates.

Technical

1. Crawl the existing site and export every URL that returns a 200. There will be pages nobody remembers.
2. Build the redirect map: every old URL to its closest new equivalent. Not to the homepage — a mass redirect to the homepage is treated as a soft 404 and the link equity is lost.
3. Record a performance baseline now. Without it, "the new site is faster" is an unverifiable claim.
4. Note every third-party script currently running: chat widgets, analytics, ad pixels, review widgets. Decide which survive.
5. List every integration: forms going to a CRM, booking systems, payment gateways, newsletter tools.
6. Check where your domain and hosting actually live, and who has access. Find this out now, not on launch day.
7. Confirm where email is hosted. Changing nameservers without moving MX records is the single most common way to take a company's email down.

During the build

1. Keep staging out of the index: robots disallow and password protection. A staging site that gets crawled becomes a duplicate-content problem.
2. Review on real devices, not just a desktop browser at full width.
3. Test every form to its failure states, not just the happy path.
4. Check keyboard navigation and run one page through a screen reader. Most accessibility problems are cheap to fix during the build and expensive afterwards.

Launch week

1. Lower the DNS TTL a day or two ahead so the switch propagates quickly.
2. Verify the redirect map on staging before cutover, not after.
3. Take a full backup of the old site and its database. Keep the old hosting alive for at least thirty days.
4. Submit the new sitemap in Search Console and request indexing for your top twenty pages.
5. Confirm analytics and conversion tracking are firing on the new site. It is easy to lose a month of data here.
6. Watch the 404 log daily for the first fortnight. It will reveal the URLs your crawl missed.
7. Expect rankings to move for a few weeks. A dip that recovers is normal; a dip that does not usually means a redirect is wrong.

The one that catches everyone

Canonical tags. We have audited sites where every page carried a canonical tag pointing at the homepage — usually because it was hardcoded into a template years earlier and never noticed. Every internal page was telling Google it was a duplicate of the homepage, and none of them ranked.

It is a one-line fix and it can be worth more than the entire redesign. Check yours before you rebuild, because if that is the actual problem, a new design will not solve it.