The short version
- Most of the work that determines whether a redesign succeeds happens before any design starts.
- Know what the current site earns before you change it. You cannot protect what you have not measured.
- Decide who writes the content on day one. It is the most common cause of a late project.
- Agree what success looks like, in numbers, before anyone draws a page.
Redesigns go wrong in predictable ways, and almost all of them are decided before the first design is presented. This is the list we work through with a client at the start, in the order it matters.
Before you brief anyone
- Write down why. In commercial terms, not “it looks dated”. What is not happening that should be?
- Define success in numbers. What should be measurably different in twelve months, and which figure shows it? If nobody can answer, the project will be judged on taste.
- Export twelve months of Search Console data. Every URL, its clicks, impressions and queries. This is your baseline and your protection.
- Find out what currently converts. Which pages produce enquiries. If you cannot tell, that is finding number one and it is worth fixing regardless.
- Agree a budget range internally. Then say it out loud to agencies. Withholding it gets you three quotes for three different projects.
- Name a decision maker. One person who can say yes. Committees add review rounds and cost.
Content, decided early
The single most common cause of a redesign running late is waiting for words and pictures.
- Who is writing the copy? Answer honestly. “We will do it internally” is right only if someone has the time booked.
- What photography exists, and is it usable? Stock imagery of people in meeting rooms actively damages credibility.
- What is being migrated, rewritten, merged or retired? Page by page, against the Search Console export.
- Who signs content off, and how long will that take? Build it into the timeline.
- Are there case studies you can name? Client permission takes weeks. Start asking now.
Client permission for names, logos and results is slow, and it is the most persuasive content on the site. Ask at the start of the project, not the week before launch.
Technical, before design starts
- List every system it must connect to. CRM, accounts, stock, booking, payment. Note which have a modern interface and which do not. This moves the price more than anything else.
- Decide the URL policy. Pages that earn keep their addresses. Everything else can move.
- Start the redirect map. A spreadsheet, old to new, built as the structure is agreed rather than the week of launch.
- Check who owns the domain and hosting. More projects than you would think stall because nobody can find the registrar login.
- Confirm code and design file ownership in writing. Before work starts.
- Plan tracking. Conversion actions, call tracking, and the source of each enquiry reaching your CRM. Built in, not bolted on.
The week before launch
- Record the baseline. Rankings, clicks, impressions, speed. Without it you cannot tell a normal settling period from a real problem.
- Redirect map complete and tested on staging, one hop, pointing at closest equivalents.
- Staging blocked from indexing, and check it has not already been crawled.
- Forms tested end to end. Submission, notification, CRM record, autoresponder. Send a real one and follow it.
- Spam protection in place. A new form with none will find the bots within days.
- Check it on a real phone on mobile data, not a desktop browser window made narrow.
Launch day
- Check for noindex. View source, search for it. This is the one that does real damage.
- Check robots.txt is not disallowing everything.
- Spot-check ten redirects by hand.
- Submit the new sitemap in Search Console.
- Confirm analytics is firing on every template, including any landing pages that use a different header.
- Test a form submission on the live site. Staging passing is not the same thing.
Planning a redesign? Send us the URL before you brief anyone. We will tell you what the current site earns, which pages are doing the work, and what needs protecting. You get the findings whether or not you work with us.
The first month after
- Week one: watch Search Console coverage for a spike in 404s. That is your redirect map telling you what it missed.
- Weeks two to four: expect movement including some downward. Do not start changing things.
- Check speed on the live site, not on staging. Real hosting, real traffic.
- Watch enquiry volume and quality, not just traffic.
- Week six: compare against the baseline properly. If you are still materially down, investigate.
Common questions
How long should we allow?
It depends far more on content and integrations than design. Decide who is writing what on day one and the timeline becomes predictable. Leave it vague and it will slip.
What if we do not have Search Console set up?
Set it up now and wait. Even a few weeks of data beats none, and going into a rebuild blind is how sites lose rankings without anyone noticing which ones.
Should we redesign and rebrand at the same time?
It is efficient if the brand work is already done. If it is not, you will end up designing pages around a logo that keeps changing. Finish the brand first.
Do we need to tell customers the site is changing?
For a public marketing site, no. For anything with logins or saved data, yes, and well in advance.
