Most redesigns start from an impression: the site “looks dated”, a competitor has just changed theirs, the team no longer likes it. The impression is often right. But it says neither what needs to change nor what must be kept at all costs.
A site that has been online for a few years is a record of your customers: the pages they read, the ones they leave, the searches that bring them in, the path they take to contact you. Redesigning without capturing that data means starting from scratch and throwing away what you knew.
What a redesign can break
A poorly prepared redesign does not just fail to improve things. It can set back what was working:
- Hard-won search rankings. URLs that change without redirects, deleted pages that used to receive traffic, rewritten content that loses what made it rank.
- Journeys that worked. A short form replaced by a “more complete” one, a phone number that vanishes from the top of the page.
- Speed. A new theme, animations, mood videos: a better-looking site can also be a slower one, and mobile pays for it first.
None of these setbacks is inevitable. All of them can be seen in advance, if the existing site has been measured.
The four measurements to take first
1. Real-world speed. Not a score taken one afternoon on a fibre connection, but field data: the Core Web Vitals Google collects from your real visitors, mobile first. They show whether the problem lies with the server, the images, third-party scripts or the structure of the pages.
2. What Google sees. The indexed pages, the searches that bring you visits, the pages that receive them. Search Console shows this for free. It is the list of what a redesign must not break.
3. The path to an enquiry. How many pages a visitor goes through before being able to contact you, book or buy, and where they give up. Your analytics tool, plus a few session recordings, is often enough to spot the breaking point.
4. What your customers tell you. The questions that keep coming up on the phone, the objections before signing, the words they use. A site that does not answer them leaves that work to your team.
What the measurements decide
Once these measurements are in, the question “should we redesign?” changes nature. It becomes a list of tasks, each with its cause and its impact.
- If the problems are on the surface — copy, hierarchy, calls to action, oversized images — targeted fixes are often enough, without starting over.
- If they reach the foundations — technical structure, an unsuitable platform, accumulated debt — a redesign is justified, and the measurements say what it must preserve.
- Either way, you know in what order to act, and you can compare quotes on the same basis.
The most common mistakes
- Measuring only once. A reading taken on a quiet day or just after an update does not tell you the trend.
- Confusing the score with the experience. A good lab score does not guarantee a fast page for your real visitors, on their phones.
- Forgetting the starting point. Without measurements before launch, there is no way to know afterwards whether the redesign improved anything.
Where to start
You can take these measurements yourself: Search Console, PageSpeed Insights and your analytics tool cover the essentials. If you would rather have an outside view first, our audit page runs a free analysis of your site in a few seconds.
And if you want to go all the way before committing a redesign budget, the Thom'Off Strategy brings together the full analysis, a consulting session and a costed roadmap.