Not Every Old Website Needs a Redesign
Since I build websites, the easiest thing would be to always propose a new one. But age isn't the problem. When I suggest a fix, a gradual improvement, a partial redesign, and when a new site really is needed.
Because I build websites, the easiest answer to any older site would always be the same: let's make a new one. It's a simple proposal, simple to offer and simple to sell.
But it isn't always the best one. Quite a few of the sites I look at don't need a new design. They need a few specific things fixed that get in the visitor's way, and to be left alone in everything else that works. A redesign without a clear reason can even make things worse: a new structure, new URLs, new texts that haven't been tested yet, and a few months during which the site has to establish itself with the search engine all over again.
So with every older site I start with a different question. Not "how old is it", but "what specifically doesn't work, and whom does it get in the way of".
Age itself isn't the problem
A site built six years ago isn't automatically bad. It may be slow, it may have awkward navigation, it may contain outdated information. Or it may be fast, clear and still say what it needs to. The year it was built tells you nothing about that.
What really interests me about an older site are a few things that can be checked in an hour. Whether it's usable on a phone without trouble. Whether a visitor quickly understands what the company does and for whom. Whether there's a clear next step, and whether it works. Whether the information is current. Whether the site sends the right signals to the search engine. Whether it's accessible to people using a keyboard or a screen reader.
The answers tell me where the problem is. Only then can I decide how big an intervention makes sense.
Where the problem is more often than in the design
When I review older sites, the same things come up again and again, and "the site is ugly" is rarely among them.
Navigation that follows the company's internal organisation rather than the visitor's questions. A mobile view that is technically responsive but has tiny text, buttons that are hard to hit and forms that don't work well on a phone. Speed, most often ruined by a few oversized images or a script nobody uses anymore. A call to action that's missing, or five of them on one page. A form that says nothing after submission, or whose messages never reach the company.
Then there's outdated information: services the company no longer offers, prices from years ago, employees who have left. Technical details that send the search engine wrong signals: duplicate titles, missing redirects, a wrong canonical. And a lack of clarity in the content, which is the hardest thing to fix with design, because the problem is what it says, not how it's laid out.
None of these problems requires a new site. Each requires a few hours or days of considered work.
The four levels I choose between
Once I know where the problem is, I use a simple scale. It isn't scientific, but it's useful, because it forces a decision about scope before we talk about price.
A fix
I choose a fix when the site basically works and has a few specific faults. A broken form, a wrong phone number, oversized images, missing page descriptions, old content on a few subpages. It's work that can be done in a day or two and has an immediately visible effect. I don't touch the design.
This is the recommendation I give most often. It doesn't sound ambitious, but the owner gets a working site for a small cost and with no risk.
Gradual improvement
I choose gradual improvement when there are several problems, but all of them can be solved within the existing design. I reorganise the navigation, rewrite the key pages, tidy up the mobile view, add clear calls to action, take care of speed and technical cleanliness. The site stays recognisable, the URLs stay, and the search engine has no reason to treat it as new.
The advantage of this approach is that we can work in steps and check each step. If the first change helps, we continue. If it doesn't, we've lost little.
A partial redesign
A partial redesign comes into play when some part of the site is designed in a way that can't be fixed without rebuilding it. A typical case is a services section conceived for a different offer, or a company presentation that has grown from three to thirty subpages without structure. I rebuild that part and leave the rest.
Here I watch the redirects. Every URL that disappears has to lead somewhere; otherwise the redesign itself creates a problem that wasn't there before.
A full redesign
A new site makes sense in a few cases. When the technical foundation can no longer be maintained safely. When the company has changed so much that the old design no longer says what the company is. When the site can't reasonably be used on a phone and fixing it would cost more than building anew. Or when the owner has no access to the system it was built in.
Then a new site isn't a luxury; it's the cheapest path. But that's a decision I make after the review, not before it.
What tends to get lost in a redesign
When we decide on a bigger intervention, I point out a few things that tend to get lost in redesigns because nobody counts them as design.
Old URLs that have had links from elsewhere for years, and impressions in the search engine. Content that was written for real customer questions and becomes a generic marketing sentence in the new version. Contact details that were correct. Structured data and metadata that someone once set up. Accessibility that was perhaps good by accident and that the new solution breaks.
A redesign that doesn't carry these forward can look better and work worse. That's why, in a redesign, I first write down what the existing site does well, and only then what we'll change.
A hypothetical case: a service company's site, five years old
To make the scale more tangible, take an invented but typical case. A small service company has a site built five years ago. The owner thinks it's outdated and asks about a redesign.
I open it on a phone. It loads slowly because of a few large photos, but is otherwise readable. The menu has seven sections, two of which lead to the same thing. The services are described in general terms, without prices and without an explanation of how things work. The contact page is clear, the form works. The company details are correct. The search engine knows the site, and for the company name it's in first place.
What would I propose? Not a new site. I'd compress the images, cut the menu to four sections, rewrite the services with the emphasis on what the customer gets and how working together goes, and add a clear enquiry prompt to every service page. That's a gradual improvement that can be done in a week or two, without changing URLs and without the risk of the site disappearing from the results for a few months.
If the same site also had a system that could no longer be updated, or a structure in which every change was a fight, the answer would be different. But I'd find that out in the review, not from the year.
What a wrong decision costs
A wrong decision in either direction has a price. A new site that wasn't needed costs money and months during which the company is busy with the site instead of with customers. It often also brings a period in which the site fluctuates in the search engine, because the addresses, structure and content all changed at once.
A fix that wasn't enough means the company keeps losing people on a site that gets in their way, and that we talk again in a year, only with less trust. So I think it matters that a review also says when a fix won't be enough. An honest assessment of scope is worth more than a favourable quote for the wrong intervention. What a price is made of and what I'd look at in a quote is covered in How Much Does a Website Cost for a Small Business.
How I decide in practice
The order I use is simple. First I open the site as a visitor and try to complete the task people come for. I've written about that in more detail in the post on reviewing a website. Then I list the obstacles and rank them by whom they obstruct and how much.
If the obstacles are isolated, I propose a fix. If they're spread out but solvable, gradual improvement. If some part is designed wrong, a partial redesign. If the foundation is wrong, a new site. With each option I also say what we won't touch and why.
That lets the owner make a decision they understand. They aren't choosing between "old" and "new", but between concrete interventions with a known scope.
The question I ask myself
Before I propose a new site, I ask myself whether I'm proposing it because the site needs it or because it's the easiest offer. That question seems fair to me, because I'm in a position where a new site is always the more interesting work for me.
The answer isn't always in favour of a fix. Sometimes a new site really is the only sensible path, and I say so just as plainly. But I want the decision to come from the review, not from a proposal template.
I make judgements like this in practice too, at GE-KO, my web development and SEO business, where a redesign starts with looking at what's worth keeping. And I'd suggest one thing to the owner of an older site before any conversation about a redesign: answer one question yourself. What exactly on the site today gets in the way of people who would like to become customers? If there's no answer, a redesign isn't the answer yet either.