There’s a conflict of interest sitting underneath this question, so let’s name it.
A rebuild is a bigger job. It pays better, it’s more interesting to work on, and it produces something worth putting in a portfolio. When you ask a developer whether your site needs replacing, you’re asking someone whose incentives point in one direction.
So here’s the test I actually use, including the cases where it costs me the work.
The question that settles most of it
Not “does it look dated.” Not “is it slow.” Those are symptoms, and symptoms mislead.
Is the underlying structure sound?
If the site is built on something maintainable, with content in sensible places and a page structure that reflects the business, then nearly every complaint you have is a fix. Ugly, slow, badly worded, clumsy on mobile: all solvable without starting again.
What makes something maintainable is mostly whether you can get at it. A theme with real templates and structured content gives you a single place to change a heading style, a spacing rule, a component. A page builder does not. The layout lives in the database as per-element settings, so there is no underlying system to edit, and the same change has to be made by hand on every page it appears.
If the structure is broken, cosmetic work is money into a hole. You’ll spend $2,500 making it better and still have fundamentally the same site, with the same problems waiting eighteen months out.
Everything below is a way of telling those two situations apart.
Fix it if any of these describe you
It’s slow, and it’s not a page builder site. A lot of slowness is dull and fixable: uncompressed images, twelve plugins doing the work of three, no caching. I’ve taken sites from eight seconds to under two without touching a single design decision. That’s days of work, not a project. Where to start when everything feels slow covers the order I work through.
The exception matters though. If the site is built in Elementor, Divi or similar, there’s a floor you cannot get under. The builder ships four times the markup the page needs, loads its own CSS and JavaScript on every request, and none of that is yours to delete. You can compress images and add caching and still land somewhere mediocre, because the weight is structural. At that point you’re paying for optimisation work that buys you a second, when a rebuild buys you five.
It looks dated, and you can actually get at the markup. On a properly built theme, a visual refresh, meaning typography, spacing, colours and a rebuilt hero, costs a fraction of a rebuild. Dated is a surface condition and the bones are fine.
On a page builder it often isn’t. Every section carries its own inherited styling, spacing lives in per-element settings rather than a system, and changing something globally means opening forty pages and repeating yourself. I’ve quoted refreshes on builder sites that came out higher than rebuilding properly, because the work is manual and there’s nothing underneath to change once.
It doesn’t generate enquiries. This one gets misdiagnosed constantly. Poor conversion is nearly always messaging and structure rather than technology: an unclear first line, contact details three scrolls down, a form asking for eight things. All fixable in a week. A rebuild that skips the diagnosis simply rebuilds the same problems in a nicer font, which is the single most expensive mistake in this whole article.
It’s awkward on mobile. If it’s responsive but clumsy, that’s CSS and a day or two. If it isn’t responsive at all, see the next section.
You find it hard to edit. Sometimes the CMS is wrong for you. More often it was set up carelessly, with fields nobody labelled and a page builder where structured content should be. Worth having someone look before you conclude the platform is at fault.
Rebuild if any of these are true
It isn’t responsive at all. A site built before mobile mattered, with a separate m-dot version or a fixed-width desktop layout, isn’t worth retrofitting. Forcing it to work costs more than building properly.
The platform is abandoned or unsafe. An unmaintained CMS, a theme whose developer vanished, a PHP version that stopped receiving security patches. At that point you’re not maintaining a website, you’re maintaining a liability with your business name on it.
Nobody can work on it. If the original developer is gone and the build is undocumented enough that every small change means two hours of archaeology, you’ll pay rebuild prices in instalments for as long as you own it.
The business has genuinely changed. New services, a different market, a real repositioning. When the site’s structure describes a business you no longer are, reorganising it honestly can cost more than starting clean.
It’s built in a page builder and you want it fast or genuinely different. Elementor, Divi and the rest are fine for standing something up quickly, and I maintain plenty of them. But there is a ceiling on both performance and design, and you hit it early. Every change fights the tooling, global changes are not really global, and the markup is not yours to trim. If the brief is “faster” or “properly ours”, you are rebuilding whether you start now or after spending a few thousand finding out.
You’re locked out of your own site. No admin access, no hosting login, a domain registered to someone who’s stopped replying. This happens more than it should, and it’s a rebuild whether or not the site is any good.
The expensive middle ground
The costly mistake isn’t picking wrong. It’s rebuilding for a reason that a rebuild can’t fix.
Plenty of rebuilds happen because someone is bored of looking at the site. That’s a genuine feeling and a terrible business case. You’ll spend the money, launch something contemporary, and in eighteen months feel exactly the same, because the thing that actually bothered you was never the design.
The other version: rebuilding to fix conversion without first working out why it doesn’t convert. New site, same vague proposition, same buried phone number, same result, larger invoice. If you can’t articulate what specifically is failing, a rebuild is a guess with a five-figure price tag.
Before either, write down what you want to be measurably different. “More enquiries” is a goal you can test against. “Modern” isn’t.
How to get an honest answer
Ask for a diagnosis before a quote. Any developer worth hiring can look at your site and tell you which category it falls into, usually inside half an hour. Someone who proposes a rebuild before asking what’s wrong has decided what they’re selling before understanding what you need.
Ask them what they’d fix if you weren’t rebuilding. This is the useful question. If there’s a credible list, you’re talking to someone thinking about your problem. If the answer is vague, you’re talking to someone thinking about their invoice.
Ask what happens to your search rankings. A rebuild without redirects from old URLs to new ones will lose you traffic that took years to earn. If they don’t raise this, raise it yourself, and watch how they answer.
What I usually find
It splits fairly cleanly, and the split is usually about what the site is built on.
Custom or properly themed sites rarely need replacing. They need a focused week: compress the images, strip the plugins nobody uses, rewrite the homepage so it says what the business does, make the phone number tappable, cut the form from nine fields to three. Not an exciting answer if you were hoping for something new, but usually the one that makes you money, and it leaves budget for being findable in the first place.
Page builder sites are the other way round. The small fixes still help, and there is a ceiling you will reach quickly. If you want fast, or you want something that looks like your business rather than a template anyone can buy, the honest answer is that you are rebuilding eventually. Spending three thousand first to find that out is the expensive route to the same place.
So work out which one you have before you decide. That question is easier to answer than “should we redesign”, and it gives you a real answer instead of a preference.