Redesign, Rebuild or Refresh? Choosing the Right Scope

“We need a new website” is the brief we hear most often, and it is right maybe a third of the time. The other two thirds need something smaller, cheaper and faster, and would have got a better outcome for it.

Three different projects hide behind the same sentence.

• A refresh changes what the pages say and how they look, on the site you already have.
• A redesign changes the design, structure and content, keeping the underlying platform.
• A rebuild replaces the technical foundation and everything on top of it.

The rule we use:

If the site works and the problem is what it says, refresh it. If the site works but you are embarrassed by it or it does not reflect what you now sell, redesign it. If the platform itself is the constraint — it is slow, insecure, unmaintainable, or nobody can edit it — rebuild.

Most people default to the most expensive of the three because it feels the most decisive. It is usually the wrong instinct.

The refresh: weeks, not months

A refresh leaves the architecture alone. You rewrite the pages that matter, replace the images, tighten the calls to action, and fix the handful of things that annoy visitors.

It is the right call more often than anyone selling websites likes to admit. A site that loads quickly, is indexed properly and has a sensible structure but describes services vaguely does not need rebuilding. It needs better words. Rebuilding it will produce a beautiful site that says the same vague things, and the enquiry rate will not move.

Choose a refresh when: the site loads acceptably, you or someone else can edit it without a developer, the structure makes sense, and your honest complaint is about the copy, the photography or the fact that nobody knows what to do when they land on it.

The upside is speed and reversibility. Change ten pages in a fortnight, see what happens, change ten more. Nothing breaks. Nothing needs redirecting. This is also the only one of the three where you can genuinely test as you go.

The limit is that a refresh cannot fix structure. If the site has no service pages, or every page is the same page with a different heading, no amount of rewriting rescues it.

The redesign: the middle option people skip

A redesign keeps the platform and replaces the layer above it. New visual design, revised page structure, rewritten content, new navigation. The database, the CMS and the hosting stay.

This is the most commonly correct answer and the one most often skipped, because “redesign” and “rebuild” get used interchangeably in conversation and then quoted as the same thing.

Choose a redesign when: the platform is fine and maintained, but the site’s structure no longer matches your business. You have added services that live nowhere sensible. The navigation has grown by accretion. The design is dated in a way that costs you credibility. Mobile was an afterthought.

The risk sits in the structure change. Moving pages means changing URLs, and changing URLs without a redirect map is the single most reliable way to lose accumulated search visibility. We have seen a well-designed new site halve a business’s organic traffic because nobody mapped the old addresses to the new ones. If you take one thing from this article, take that.

Before any redesign, list every URL that currently receives traffic or has links pointing at it, and decide where each one goes. It is dull work and it is not optional.

The rebuild: justified less often than it is sold

A rebuild replaces the technical foundation. New platform or a clean install of the same one, new templates, new integrations, content migrated or rewritten.

Choose a rebuild when one of these is true:

• The site runs on software that is no longer maintained or updated, which makes it a security liability rather than an inconvenience.
• Nobody can edit it without a developer, and that dependency is costing you more each year than the rebuild would.
• Performance problems are structural rather than cosmetic — the theme itself is the reason it is slow, and no amount of image compression fixes it.
• What you need next is impossible on the current setup. Ecommerce, bookings, a membership area, a second language.
• It was built by someone who has vanished and nobody has the access.

Note what is not on that list: “it looks old”, “we don’t like it”, “it’s been five years”. None of those are technical problems, and none of them require replacing the foundation.

A rebuild is the longest and most expensive of the three, and it carries the most risk. Everything that could go wrong in a redesign can go wrong in a rebuild, plus data migration, plus the possibility that the new platform does not do something the old one quietly did.

The diagnosis questions

Run through these before anyone quotes you anything.

1. Can you edit a page yourself today? No means the technical layer is a constraint.
2. Does the site load in a couple of seconds on a phone on mobile data? Test it on an actual phone, not your office broadband.
3. When did the platform last get an update? If nobody knows, that is your answer.
4. Do your services each have a page that explains them properly? If not, that is a content problem, not a technical one.
5. What do the analytics say people actually do? If most visitors land and leave from one page, the problem is that page.
6. What exactly is the complaint? “It’s ugly” is a design brief. “Customers ring to ask things the site should answer” is a content brief. “It goes down” is a rebuild brief.

The uncomfortable advice

If your site gets very little traffic, redesigning it will not fix that. A new site with no visitors is the same problem in better clothing, and it is a costly way to avoid confronting the real question, which is how anyone finds you at all.

We turn down rebuild enquiries fairly regularly for that reason, and suggest a refresh plus a plan for visibility instead. It is a smaller invoice and a better outcome, and we would rather have the second one.

There is also a legitimate answer of “do nothing yet”. If the site is working and nothing on the diagnosis list is failing, spend the money somewhere it earns.

If you want a view on which of the three your site needs, describe the complaint rather than the solution when you get in touch. Our website development page sets out how we scope this kind of work.