Rebuild or refresh? How to tell what your website needs

Someone decides the website is the problem and the answer is a rebuild. Sometimes that's right. Often it's an expensive way to avoid a smaller, harder conversation.

Here's how to tell.

Refresh when the bones are sound

If these are true, you probably don't need a rebuild:

• It loads reasonably fast, or would with compressed images
• It works properly on a phone
• You or someone on your team can edit content without a developer
• It's on a platform with active support
• The structure roughly matches what you sell
• It's secure and up to date

If so, your problems are almost certainly message, content and conversion — none of which require new code.

A refresh means rewriting the headline and key pages, replacing weak imagery, splitting a bloated services page, adding proof, and fixing performance. Costs a fraction of a rebuild and often produces more improvement, because the actual problem was what the site said.

Rebuild when the foundation is against you

Some things can't be fixed by editing:

You can't update it. Every change needs a developer who may not be reachable. This kills more sites than bad design — nothing gets updated, so nothing improves.

It's not really responsive. Some older sites have a separate mobile version, or none. Retrofitting is often more work than starting over.

The platform is abandoned. Unsupported CMS, discontinued builder, custom system nobody understands. A security incident waiting to happen.

It's fundamentally slow. Not image-slow — architecture-slow. Bloated themes with a decade of plugins can't be optimised into being fast.

The structure can't hold what you sell. If you now offer six services and the site was built for one, and there's no sane way to add pages.

You don't control it. Agency owns the domain or hosting. Fix this immediately regardless of anything else.

The honest diagnostic

Ask what specifically is failing.

"It looks dated" — refresh. Design is the cheapest part to change.

"It doesn't bring in enquiries" — probably not a rebuild. That's usually message and conversion. A new site with the same weak headline performs the same.

"I can't add a page without paying someone" — rebuild. That constraint compounds forever.

"It's slow" — measure first. Usually images, which is a refresh. Occasionally architecture, which is a rebuild.

"It doesn't work on phones" — test it properly. Broken layout is a refresh; no responsive foundation is a rebuild.

The order that avoids waste

Even if you're rebuilding, do the thinking first.

Sort out positioning, messaging and structure before anyone designs anything. A rebuild that carries over the old unclear message produces a better-looking site with the same problem — and you've spent the budget.

The most expensive rebuilds we see are the ones where nobody asked what the site should say until the design was already approved.

A cheap experiment

Before committing, rewrite your homepage copy on the existing site. Just the words — headline, opening, calls to action.

Give it a month.

If enquiries improve meaningfully, your problem was the message and a rebuild would have been an expensive way to change the words. If nothing moves and the site is genuinely constrained, you've lost a month and gained real evidence for the decision.