Most people asking for a redesign are describing a rebuild, and most people quoted for a rebuild only needed three things fixed. The words get used interchangeably, and the difference is usually several thousand dollars.
Worth being clear about the terms first:
- A redesign changes how the site looks and reads. Same platform, same structure underneath, new surface.
- A rebuild replaces the foundation. New platform or new code, usually new structure, and the design comes along with it.
The distinction matters because a redesign applied to a bad foundation is money spent decorating a problem you’ll still have afterwards.
Signs a redesign is enough
- The site does convert — people find you and get in touch — it just looks like it was designed in a different decade.
- Your staff can update content without breaking things.
- It loads reasonably fast, and it’s fine on a phone.
- The platform is current and supported.
- The structure makes sense; you’re not fighting the navigation to explain your business.
If most of that describes your site, a redesign is the right purchase. Don’t let anyone talk you into more.
Signs you need a rebuild
- Nobody can change it without something else breaking. The clearest signal there is.
- The platform or language version is unsupported. No security patches, and your host will eventually force the issue at a time of their choosing rather than yours.
- It’s slow for structural reasons — twenty plugins doing overlapping jobs, a database nobody has indexed, a page-builder generating a megabyte of CSS.
- It doesn’t really work on a phone. Not “looks a bit cramped” — genuinely doesn’t work.
- It has been compromised, more than once. Reinfection usually means the original entry point is still there.
- Nobody knows how it works. The person who built it is gone and the code is undocumented.
The test that actually settles it
Ask this: how much of the current system would still be there after the redesign?
If the answer is “the platform, the database, the integrations, the structure — we’re changing colors, type, and layout,” that’s a redesign and it’ll be good value.
If the answer is “honestly, not much,” you are rebuilding whether or not anyone calls it that. And in that case, doing a redesign first means paying twice: once to make the old thing look better, and again in eighteen months when the underlying problems surface anyway.
The cost comparison people run is redesign-now versus rebuild-now. The comparison that matters is redesign-now-plus-rebuild-later versus rebuild-now.
The third option nobody offers you
Sometimes the right answer is neither, and this is the one we end up recommending more than either of the others.
A site that converts poorly often has three or four specific, fixable problems: the phone number isn’t visible on a phone, the contact form is broken, the homepage asks visitors to choose between eight things, the top three pages take six seconds to load. Fixing those is a fraction of the cost of a redesign and frequently produces more revenue, because you’re repairing the conversion path rather than repainting it.
Nobody quotes this because it’s a smaller invoice. It is often the honest answer.
Related, and worth saying: if you’re not getting found at all, a new design won’t fix that. Neither a redesign nor a rebuild moves you up in search results by itself. For a local business, that’s usually Google Business Profile and reviews, and it’s cheaper than either.
Whichever you do: keep your URLs
This one gets skipped constantly and it is the most avoidable expensive mistake in the whole category.
Every page on your existing site has an address. Some of those addresses are in Google’s index, in other people’s links, in emails, in printed material. If a rebuild changes them without redirecting the old ones to the new ones, you throw away whatever search visibility you had and you break every external link pointing at you.
The fix is a redirect map: every old address, every new address, and a permanent redirect between them. It is tedious and completely mechanical, which is exactly why it gets skipped.
We rebuilt this site recently and it had fifty-seven pages. Every one of them is either still at its original address or permanently redirected to a deliberate destination, and there’s a script that fails the build if any path is unaccounted for. That’s the standard to hold whoever does your rebuild to — not a promise that they’ll be careful, but a list you can check.
What we’d tell you on a call
We’d ask what the site is failing to do, not what you dislike about it. Those are frequently different problems, and only the first one costs you money.
Then we’d tell you which of the three you need, including when the answer is “spend a tenth of this budget on four fixes and put the rest into getting found.” That’s the less profitable answer, which is why it’s worth asking for.
If you want that conversation, tell us what the site isn’t doing and we’ll give you a straight read on which one it is.