Should I Redesign My Website or Fix the One I Have?

The redesign conversation usually starts with a feeling rather than a finding. The site looks dated next to a competitor's. The owner is embarrassed to send the link. Someone on the team has a friend who builds sites. None of those is a reason, and all of them lead to the same expensive place: a new site that looks better, launches to applause, and produces the same number of calls as the old one, because the thing that was limiting calls was never the design.

What follows is how I separate the two cases. It is deliberately biased toward fixing, because the fixes are cheap to try and the rebuild is not, and because a rebuild that was truly needed will still be needed after the fixes and will be better informed by them.

Five questions that decide it

1. Does it work on a phone? Not "is it responsive," which nearly every site built in the last decade is. Can a stranger find a product or service, get to the point of buying or booking, and finish, on a phone, on cellular, without pinching, squinting, or giving up? Contentsquare's 2026 benchmark puts mobile at about 70 percent of visits, with desktop converting 74 percent higher than mobile web, so this is where most visitors are and where most of them leave. If the answer is "mostly, with a few rough spots," that is fixes. If the answer is that the site fundamentally does not work on a phone, that is the strongest case for a rebuild on this list.

2. Do the pages match the business? Companies change faster than their websites. Services get added and dropped, a product line becomes the main line, a new market opens. When the site's structure describes a business that no longer exists, adding pages to it makes a maze. When the structure is right and the pages are merely thin or dated, that is copy and content work on the existing frame.

3. Is the platform the limit? A site on a builder that cannot take a proper checkout, cannot be edited without a developer, loads slowly no matter what is removed, or cannot be connected to the tools the business now uses has hit a ceiling that no fix will raise. That is a rebuild reason. A site on a capable platform that has simply been neglected is not.

4. Is anything ranking that a rebuild could lose? Open Search Console and look at which pages bring visitors from search. Those rankings belong to specific URLs and specific copy. A rebuild that changes URLs, drops pages, or rewrites the content that matched the searches will lose them, and rebuilding them takes months. If the site has real search traffic, a rebuild needs a redirect plan and a page-by-page content plan before anything else, and "fix" becomes more attractive still.

5. Can you name the three biggest problems? If the answer is "the checkout is four steps, the forms are too long, and the homepage takes six seconds on a phone," you have a fix list, and it can be done in weeks. If the answer is "it just feels wrong," the honest next step is not a rebuild; it is finding out what is wrong, because a redesign that starts from a feeling reproduces the problems in a new coat of paint. The free 12-point audit and the checkout leak calculator exist to turn the feeling into a list with numbers attached.

When fixing wins

Most of the time. The problems that cost small businesses money on their websites are, in rough order of how often I see them: nothing is being measured, so nobody knows what works; the checkout or the form asks too much; the page is slow on a phone; the page people land on from an ad or a search does not match what they clicked; and the copy does not say what it costs or what happens next. Every one of those is a targeted fix. None of them is solved by a redesign as such, and a redesign done without addressing them carries them straight into the new site. The six-cause article on traffic but no sales works through them in order.

There is also a measurement argument. Three fixes made one at a time can each be judged: the mobile checkout changed on a Tuesday, and orders from phones moved or did not. A redesign changes everything at once, and if the number goes up nobody knows why, and if it goes down nobody knows which part to undo.

When a rebuild is right

Sometimes it is, and pretending otherwise would be its own kind of sales pitch. A rebuild is the right call when the platform is the ceiling, when the structure describes a business that no longer exists, when the site cannot be made to work on a phone without starting over, or when fixes have been tried in earnest and the underlying build makes each one cost more than it should. It is also right, occasionally, for a business whose entire market is visual and whose site is visibly a generation behind, provided the questions above have been asked first and the answers still point the same way.

When it is right, do it in a way that keeps what works. Keep every URL that ranks or redirect it one to one. Keep the pages that bring search traffic and improve them rather than replacing them. Fix tracking before launch so the new site can be judged from day one. Set up the new site to be edited without a developer, because the next set of fixes should not require another rebuild. And check Search Console the week it goes live, not the quarter after.

A reasonable way to decide

Run the five questions. If the platform is not the limit, the structure roughly fits, and the site can be made to work on a phone, write down the three biggest problems and fix those, one at a time, measuring each. Give it two months. If the number moved, keep going. If the fixes were made in earnest and the site still cannot do what the business needs, you now have a rebuild brief written in specifics rather than feelings, and a rebuild done from that brief is a very different project from one that started with "it looks dated."

If you would like the five questions answered by someone who builds sites and still says "fix it" most of the time, that is what the free consultation is for.