Do I Need a Website Redesign, or Just a Few Fixes?

Every few years, someone tells you your website looks dated. Sometimes it is a designer with a proposal, sometimes it is your nephew, sometimes it is you at eleven at night comparing your site to a competitor's. The suggestion is always the same: start again. New look, new structure, new everything. And it does feel like the honest answer, because if enquiries are flat and the site looks tired, the two must be connected.

They often are not. A redesign is one of the few decisions a small business makes where the cost is certain and the benefit is entirely a guess. What follows is a way to check, before you commit, whether the design is genuinely what is holding you back — or whether three afternoons of unglamorous fixes would do more. The evidence is already in your own analytics. You just need to know which numbers separate the two cases.

The money is the obvious cost, but it is rarely the largest one. A rebuild takes your attention for weeks. You write copy, you chase a developer, you approve mockups, you argue about a photograph of the shopfront. During that time nothing else about your marketing improves. Then, when the new site goes live, the useful history you had — which pages brought enquiries, which search terms landed people on you — gets scrambled, because the page addresses have changed and the old records no longer line up with the new ones.

There is also a quieter risk. A redesign changes everything at once, which means that if enquiries go up you will not know why, and if they go down you will not know why either. You have replaced a site you understood with one you do not, and you have given up the ability to learn anything from the change. That is a fair trade if the old site was genuinely broken. It is a poor one if the problem was a phone number that was hard to find on a mobile.

None of this means redesigns are always wrong. It means the burden of proof should be high, and it should be met with evidence rather than a feeling.

There are a handful of situations where patching will not get you there, and each of them shows up in the data or in five minutes of honest looking.

The first is a structural mobile failure. Not "it looks a bit cramped" — genuinely unusable. If most of your visitors arrive on a phone, and the phone version of a page has a much worse outcome than the desktop version of the same page, that gap is design, not copy. Compare like with like: the same page, same month, phone against laptop. A small difference is normal. A large and consistent one across every page is a rebuild signal.

The second is that you cannot change anything without paying someone. If adding a paragraph to your services page requires an email to a developer who replies in a fortnight, your site is a liability regardless of how it looks. A site you can edit yourself is worth more than a site that is prettier and frozen.

The third is that the business has genuinely changed. The solicitor who now does two-thirds conveyancing but has a site built around family law is not going to fix that with better buttons. The same goes for a plumber who has quietly become a bathroom fitter. When the site describes a business you no longer run, no amount of tuning will help, because every page is answering the wrong question.

The fourth is technical rot: the platform is no longer supported, the security certificate keeps lapsing, pages take many seconds to load and nobody can work out why. If the foundations are failing, you are not redesigning, you are rebuilding, and that is a different and more justifiable decision.

Now the more common case. People are arriving, they are reading, and something specific is stopping them at a specific point. That is not a design problem. That is a leak, and leaks have locations.

Look at where people leave. The exit page is simply the last page someone saw before they closed the tab or went elsewhere — and it is a far more useful number than an overall site average, because it tells you the room they walked out of rather than the fact that they walked out. If your exit pages show that visitors are happily moving through your services page and then leaving on the contact page, the design of your homepage is irrelevant. Something on the contact page is the issue: a form that asks too much, a phone number that is an image rather than a tappable link, or a map that takes forever to load.

Look also at how far down people read. If visitors are scrolling to the bottom of a page and then leaving, they read the whole thing and were not persuaded — a copy problem. If they stop a third of the way down every time, they lost interest early, which is usually a headline or a first paragraph that talks about you rather than them. Those two problems look identical in a summary report and completely different when you check scroll depth page by page.

And look at which pages actually carry the business. Most small sites have two or three pages doing nearly all the work and a dozen that almost nobody visits. If your top pages report shows that four pages account for the overwhelming majority of visits, a full redesign spends most of its budget on pages nobody looks at. Fixing those four is a much better use of a fortnight.

These are dull, which is exactly why they get skipped in favour of something that feels like progress. They also tend to move the needle more reliably than a new colour palette.

Speed is the first. Not "make the site fast" in the abstract, but find the one page that is slow and find the one thing making it slow, which is almost always an enormous uncompressed photograph or a third-party widget nobody remembers installing. A booking calendar, a chat bubble, a review carousel — each adds load time, and on a phone on a patchy signal, that is where the visitor goes. Tracking page speed over time rather than testing once tells you whether something has got worse and roughly when, which usually points straight at what changed.

Broken links are the second. Every site accumulates them: a supplier who rebranded, an old PDF price list, a blog post that linked to a page you deleted. They cost you nothing to find and a few minutes to fix, and a dead link on a page where someone was ready to act is a genuinely lost enquiry.

Then there is the contact route. Count the taps between the homepage and someone actually reaching you. If it is more than two on a phone, that is your fix. Put the phone number in the header as a tappable link. Cut the enquiry form to name, contact detail, and a message box — every extra field is another reason to abandon it. Say what happens after they send it, because "we usually reply the same working day" removes a real hesitation.

Finally, the words above the fold on your two busiest pages. What you do, where you do it, and what it costs or roughly costs. An astonishing number of sites make a visitor scroll to find out whether the business even covers their area. That is a twenty-minute fix that no redesign brief ever includes.

Here is the sequence. It takes about an hour if your analytics are already running.

Start with volume. If fewer than a few hundred people a month reach your site, your problem is almost certainly not the site — it is that nobody is arriving. Redesigning it will change nothing, because a better shop window does not help on an empty street. Fix visibility first.

If people are arriving, split them by device and compare the same page across phone and laptop. Then find the page where the most people leave and ask what that page is asking of them. Then check how quickly your three busiest pages load on a phone. Then, and only then, ask whether the thing you found is a page-level fix or a whole-site failure. In practice — and this is a rough impression from looking at a lot of small business sites, not a statistic — most owners who arrive convinced they need a rebuild find two or three specific pages doing the damage.

If you would rather not do that manually, our free site report card runs the mechanical parts of that check for you: speed, broken links, mobile behaviour, and which pages are carrying the site. It will not tell you whether your brand looks tired. It will tell you whether the tiredness is costing you anything measurable, which is the question that actually determines the answer.

Sometimes the evidence really does say rebuild. In that case, spend the first week not designing anything. Write down which pages currently bring enquiries, which search terms bring people in, and what the current outcome rate is on your two or three most important pages. Without that baseline you will never be able to say whether the new site is better, and you will spend the next year defending a decision on instinct.

Then insist that every old page address either survives or forwards to its nearest equivalent. This is the single most common way a redesign loses traffic: pages that Google had indexed for years quietly vanish, and search visits fall for months while the new pages earn their place back. Ask your designer directly how the old addresses will be handled, and get the answer before work starts.

Keep the things that worked. If a slightly ugly testimonial block sits just above your contact form and the numbers say people who scroll past it get in touch more often, keep it, ugly and all. Designers are trained to make things coherent. Coherence is not the same as effective, and you are the one who has the evidence.

Whichever route you take, change things in a way you can learn from. If you make five fixes in one afternoon, you will never know which one mattered. Make one, note the date, wait a fortnight, and look. Marking the date directly in your analytics with a short note — "cut contact form to three fields" — means that in six months you can look at a change in the line and know exactly what caused it.

Judge it on the right measure. Not visits, which move for reasons that have nothing to do with you. Judge it on the proportion of people who reach a page and then do the thing you wanted — ring, book, email, buy. That figure moving from a low number to a slightly less low number on your busiest page is worth more than a site that wins compliments.

This is the part where a plain analytics dashboard leaves you on your own, and it is why Crafty Meerkat sends a short weekly note with the specific things worth changing on your site, ranked, in order. If you would like to see what that would say about your site before deciding anything, book a demo and we will walk through your actual numbers rather than a sample account.