Updating a website or building a new one: how to tell the difference
Your site was built seven years ago. Text jumps around on a phone, you can’t change a headline without emailing someone, and the contact form stopped working after the last update. The question is simple: is this still fixable, or is it time to start over?
The answer comes down to a handful of things you can check in an afternoon. Not a gut feeling, and not how the site looks. Here’s what we go through when someone brings us an old website, and how that turns into a decision.
Why websites age faster than you’d expect
A website isn’t a piece of furniture that sits there unchanged. Everything around it keeps moving. Browsers ship new versions, hosts push PHP upgrades, payment providers change their APIs, Google adjusts what it rewards. A site nobody touches is, three years later, running in an environment it was never built for.
Then there’s the other kind of ageing. Your business changed. Different services, different clients, a different story to tell. The site can be technically spotless and still describe a company that no longer exists.
These two get confused constantly. Someone says “we need a new website” when what they actually need is new copy and better photos. And the reverse happens too: a site looks perfectly decent while, under the hood, every change costs three times what it should.
Six signs your site needs attention
1. It works worse on a phone than on a desktop
More than half your visitors arrive on a phone. If they have to pinch to zoom, if buttons sit too close together, if the menu opens every other tap, a good share of them simply leave. Google knows this and judges your site by its mobile version first.
2. Only a developer can change the text
If updating a phone number in the footer means emailing someone and waiting two days, the site isn’t doing its job. Copy, images, prices and news should be something you handle yourself in a few minutes.
3. It takes longer than three seconds to load
You don’t need to guess at this — put the address into PageSpeed Insights and read the numbers. A slow site costs you twice: some visitors never wait for it to appear, and Google ranks it below faster competitors.
4. You’re afraid to press “update”
When dozens of pending updates pile up in the admin panel because everything broke the last time, that’s no longer a cosmetic problem. An old, unpatched system is the single most common way malicious code gets into a website.
5. Google shows the site, but nobody clicks
Search Console might report thousands of impressions against a handful of clicks. That usually means your page titles and descriptions don’t match what people are actually searching for, or you’re sitting on page two. Content fixes this. A new design doesn’t.
6. It doesn’t say what you actually do
This is the subtlest one, and the hardest to spot yourself, because you’ve stopped reading your own site years ago. Good test: hand it to someone who knows nothing about your business and ask what they understood in thirty seconds. The answer is often not what you expected.
When updating is enough
Updating means the foundation stays and the pieces on top change: design, copy, structure, speed, specific features. When does that make sense?
The system is still supported. A WordPress install that gets its updates, or a Laravel version still receiving patches, is a perfectly reasonable base to work on.
The code isn’t held together with tape. If the site follows normal theme and plugin conventions without too much improvisation, working on it holds no surprises.
The structure broadly fits. The page tree matches your services, URLs make sense, and there’s no leftover debris from three previous redesigns.
You know what’s wrong. When the problem is specific, the work has clear edges and a clear price. When it’s vague, the scope keeps growing and nobody ends up happy.
In cases like these, an update delivers most of the benefit for a fraction of the cost. Typically that means a new design on the existing system, rewritten copy, performance work, a proper mobile experience and deeper service pages. It takes anywhere from a few weeks to a couple of months, and the site stays live throughout.
When a rebuild costs less
Sometimes an update only looks cheaper until someone counts the hours actually involved. A rebuild is worth considering when you recognise two or more of these.
The system is no longer supported. An old Joomla install, Drupal 7, Magento 1, or bespoke PHP with no documentation. Every change becomes archaeology, and nobody is patching the security holes any more.
Each fix creates a new problem. That’s the signature of code layered on code until the logic disappeared. Continuing to patch it is pouring money into a hole.
The structure no longer matches the business. When a site was built for selling products and the company now sells services, you end up rebuilding navigation, pages, URLs and copy. That’s a rebuild wearing a different name.
You need things the current system can’t carry. A customer portal, an accounting integration, multiple languages, serious search. You can bolt these onto an old base, but what you get is fragile.
One thing that gets overlooked far too often: when you rebuild, the old site has to be migrated properly. If URLs change without redirects, Google loses everything the site had built up and rankings fall off a cliff. Done well, traffic usually rises after a rebuild. Done badly, it takes six months to climb back. This isn’t a detail — it’s a core part of the project.
What drives the cost
Pricing varies more than most people expect, and it follows scope rather than page count.
Targeted fixes are the cheapest route: sorting out the mobile experience, performance work, rewriting copy, adding a few pages. This works when the problem is specific and contained.
A redesign on the existing system sits in the middle. Look and structure change, the platform stays. The result is often indistinguishable from a new site, at a noticeably lower cost.
A full rebuild is the largest number, and it grows with integrations, languages and custom functionality. Business systems and customer portals are a different category again.
Worth adding to whatever figure you’re quoted: the cost of the decisions themselves. Writing copy, sourcing photography, reviewing drafts. This is almost always what determines whether a project runs to schedule.
Four questions to ask yourself
If you need an answer quickly, work through these in order.
Is the system still supported? If not, there’s nothing to weigh up.
Have you paid to fix the same problem more than once in the past year? If yes, money is already going into propping up something that doesn’t work.
Does the structure match what you sell today? If not, a new design won’t solve it.
Does Search Console show that Google sees the site at all? If impressions are near zero, the problem is content, and it won’t disappear with a redesign.
Three “no” answers out of four point fairly clearly to a rebuild. One or two, and an update will probably do.
How we approach it
Before recommending anything, we run a technical audit: the platform and its versions, the state of the code, performance, mobile, security, integrations, and how the site looks to Google. That turns into a short document with findings and two paths — what an update would give you, what a rebuild would give you, with rough costs and timelines for each.
The decision is yours. There’s no sense in selling a rebuild where a tidy-up would do, because long-term clients only come from honest proposals. We work with WordPress, WooCommerce, Shopware and Laravel, and after twenty-seven years we’ve taken over enough sites to spot where the surprises tend to hide.
If your site works and you’d like it to stay that way, website maintenance covers the updates, security and monitoring that keep this question from coming back in three years. Online stores have their own plans, because there are more moving parts.
And if it’s already clear you need to start from a blank page, website development is where to look next.
