"The website is four years old now." It's how almost every conversation about a redesign starts, and it's the worst possible reason to spend the money. A website doesn't expire by date. It expires when it stops doing what it's meant to do, and that can happen at six months or never.
The problem is that a full redesign costs what it costs, and half the time the real problem could have been fixed by changing three things. Before rebuilding, it's worth telling the difference.
Which signals do justify a redesign?
Changing something small takes days, not minutes. If updating a price, uploading a photo or posting your holiday closing means waiting for someone to have a free slot, that isn't a design problem: it's a structural one, and changing the colours won't fix it.
Mobile works like a patch job. Not "looks a bit off": actually works badly. Forms you can't fill in with your thumb, menus that cover the content, a call button that doesn't call. Today mobile isn't a secondary version of your website — for a local business it's the real front door: people arrive from the map, from a review, from a link, almost never by typing your address into a computer. If that experience fails, you're not losing one odd visit — you're losing almost all of them.
The website doesn't say who you are today. You started out as a graphic designer and today 70% of your work is video. You started with one focus and today you do something else, or tell it differently. The website still tells the old story. This isn't cosmetic: Google ranks what it reads, and people believe what they see.
Performance is genuinely poor. You don't need a report to notice it: the page takes to load, images jump around, you tap a button and nothing happens for a long second. That's not a technical detail — it's the feeling whoever visits is left with, and that feeling turns straight into how much they trust your business. If your site is a WordPress with a dozen plugins stacked on top and you notice that sluggishness yourself browsing on your phone, you don't need to measure much more to know the redesign pays for itself.
And there's one signal no tool will show you that's worth more than all of them: people land on your website and then phone you to ask something the website already answers. That means the information is there but can't be found. That's an architecture problem, and that one really is fixed by rebuilding.
What is NOT a reason to rebuild?
A typeface that's aged, a competitor launching a flashier site, or simply being tired of looking at your own aren't, on their own, reason enough for a redesign — usually a facelift covers it. Wanting a new logo isn't either: that's branding, and it can be applied to the current site. And if what's actually failing is that you don't show up on Google, the problem is almost always the content or your Google profile, not the design.
Full redesign or piece by piece?
It depends where the fault is:
| Situation | What to do |
|---|---|
| Dated look, everything else fine | Visual redesign on the existing structure |
| Outdated content, solid site | Rewrite pages and add the missing ones |
| You can't edit it yourself | Change the platform: here, yes, rebuild |
| Fails on mobile and on performance | Rebuild, and measure before and after this time |
| You sell different things than three years ago | Rebuild, starting with the structure, not the design |
In short
Rebuild your website if you can't manage it yourself, if it fails on mobile, if it no longer says what you do today, or if people can't find what's inside it. Don't rebuild it just because it's a few years old.
And if you're not sure which of the two you're in, there's a simple way to find out before asking anyone for a quote: look at how many visits you get, how many of those turn into a call or a form, and how the site actually feels browsing it yourself on your phone. With that on the table, the answer usually gets pretty clear.
At GMB Horizon we always start by looking at that before proposing anything: web design. If you'd like us to take a look at yours, no strings attached, get in touch.