Rebuild or redesign? How to decide what your website actually needs

The request usually arrives as 'we need a new website'. When we dig in, the real problem is one of three things: the site is slow, nobody can update it, or it does not do what the business now needs. Each of those has a different answer.

Start with a short audit

Before recommending anything we look at four areas: performance (real Core Web Vitals data, not a single lab score), the content workflow (who edits what, and how painful it is), the codebase (framework age, dependency health, test coverage) and the integrations the site depends on.

When a refactor is enough

• The framework is still supported and the code is readable
• Performance problems come from images, scripts or hosting rather than architecture
• The content model fits the business, even if the editing interface is dated

In these cases a focused refactor, often a few weeks, gets most of the benefit at a fraction of the cost of a rebuild.

When a rebuild is the honest answer

If the site cannot support the next two years of the business, a rebuild is usually cheaper than a series of patches. Signs include a framework that is no longer maintained, a content model that forces workarounds for every new page type, or integrations held together by manual steps.

The most expensive website is the one you rebuild twice because the first rebuild solved the wrong problem.

Either way, keep what works

Content, URL structure and search rankings are assets. A rebuild should carry them over deliberately, with redirects and a content migration plan, not start from a blank page.