Do I need a new website, or can mine be fixed?

Usually it can be fixed. That is the ordinary finding, not a polite one. Most sites that are not performing have specific faults, and specific faults can be repaired for far less than £3,500.

The three things that make a site worth replacing

A new website earns its cost when the structure itself is the limit, not when the site is merely dated. There are three of those, and you can test all three yourself in an afternoon.

It cannot be made fast. Not slow today, but slow in a way that cannot be undone. Run your home page through PageSpeed Insights on mobile. Then look at what is loading: if the weight is images and a chat widget, that is fixable in a week. If the weight is the page builder that every page is made of, it is not, because removing it means remaking every page anyway.

Machines cannot read it cleanly. Ask an AI assistant to recommend a business like yours in your town, and see whether it names you. Then look at why. If your services are described in text, an assistant can quote them. If they live inside images, sliders or tabs that only open on a click, it cannot. That is a structural problem when the whole site is built that way.

You cannot add a page without paying someone. If publishing a new service page means a developer invoice every time, the site is a cost that grows. That is the one owners underestimate most, because it does not show up until the second year.

Does the platform matter

Less than the arguments suggest. WordPress, Astro, Next.js: the platform is rarely the reason a site is slow or invisible.

A slow WordPress site is almost always slow because of what was added to it. Page builders, a dozen plugins, a theme that ships code for features you do not use. The same site remade carefully on the same platform is usually fast. We have found more sites held back by their plugins than by their platform.

Where the platform does matter is what it costs to keep. A site that serves plain pages needs little maintenance and stays fast on its own. A site assembled from plugins needs someone watching it, because each plugin is another thing that updates, breaks or gets abandoned. That is a running cost argument, not a search one.

So: moving platforms will not fix a site that is badly structured, and staying put will not hold back a site that is well structured.

Fix or replace, by symptom

If you do replace it, what protects your rankings

The fear here is real and worth naming: people lose visibility in a move, and when it happens it is usually months before anyone notices. It is avoidable. Four things do the avoiding, and you can ask any supplier whether they are doing them.

Every old address points at its new one. Each page that existed before gets a permanent redirect to the page that replaces it, one to one. Not everything pointed at the home page, which is the shortcut that loses the most. If a page has no replacement, the nearest honest match is better than nothing.

The content comes across, not a summary of it. A page that ranked did so because of what was on it. Cutting three quarters of the words for a cleaner look removes the reason it ranked. Say it better by all means, but say as much.

It launches faster than it was, not slower. A new site that is prettier and heavier is a downgrade Google can measure. This is the one most often missed, because the new site feels faster on the office broadband.

Your name, address and phone number stay identical. Same wording on the site and on your Google Business Profile, down to the abbreviations. Changing the address or phone in several places at once is a common way to have a profile suspended, and getting one back takes weeks.

None of this is exotic. It is the checklist, and the reason to ask for it up front is that it costs nothing to do properly and a great deal to repair.

When we cannot tell from the outside

Sometimes the answer is genuinely unclear until someone reads the code and the Search Console data together. That is what the SEO audit is for. It ends with a written recommendation, and that recommendation is allowed to be “keep this site”.

Several audits have ended exactly there. That is a finished job, not a failed sale.