A redesign that does not throw away what already works

Before anything is drawn, the current site gets read the way a visitor and a crawler read it: which pages get traffic, which ones convert, where people leave, and which URLs other sites link to.

That list is the constraint the redesign is built inside. A site that looks better and performs worse is a very expensive way to lose ground, and it is the normal outcome when the new design is briefed only on how the old one looks.

The single most common way a redesign costs money is URLs. Pages move, slugs get tidied, a section is merged into another, and everything that pointed at the old address lands on a 404 — including the search results that were sending you customers.

Every existing URL is mapped to its destination on the new site before launch, and the ones with nothing equivalent get a deliberate decision rather than a default. This is unglamorous and it is the part that protects the investment.

A full-site relaunch on one date is the riskiest version of this work. Where the site allows it, the redesign goes out in stages — the templates that carry the most traffic first — so problems surface on one section rather than on all of them at once.

It also means value arrives earlier. The pages that matter get better in month one instead of everything getting better in month five.