The reason this decision goes wrong so reliably is that a website's problems live in different layers, and the layers fail at different speeds, cost wildly different amounts to fix — and are visible to completely different people. The director sees the design. The office sees the outdated staff page. Nobody sees the plugin that has not been patched since 2023. So the loudest complaint sets the budget, and the loudest complaint is almost never the most expensive problem.
The four layers.
Content — what the site says
Stale projects, old staff, a missing service, the wrong phone number. Embarrassing, damaging to trust — and the cheapest layer to fix. No rebuild has ever been required to update a paragraph.
Presentation — how it looks and reads
The dated design, the tiny text, the layout that falls apart on a phone. This is the layer that triggers the "rebuild" conversation — and it is a redesign problem, not a rebuild problem, unless deeper layers are also failing.
Platform — what it runs on
The unmaintained WordPress install, the plugin stack nobody dares update, the page builder whose company is winding down, hosting held by someone who no longer answers. Invisible from the outside while the site "works" — and no amount of content editing or redesign cures it.
Foundation — whether it fits the business
The company changed and the site's structure did not: services that no longer exist, no place for what you actually sell now, information organised for a business you stopped being. When this layer is wrong, the other three inherit the problem.
The verdict follows the layer.
Repair when the platform is sound and the complaints live in content or presentation — targeted work, priced in days not months, and the most under-sold option in the market because nobody earns much recommending it. Migrate when it is the reverse: the content and structure still serve the business, but the platform underneath is rotten or hostage — unmaintained WordPress being the classic case. Moving the good content onto a sound foundation preserves what works and abandons what does not. Rebuild when the foundation itself is wrong — the structure cannot say what the business now needs said — or when a takeover found nothing recoverable. It is the right answer far less often than it is the reflex answer.
And the underrated fourth verdict: leave it alone. A plain, dated site that ranks for your company name, loads fast and quietly produces enquiries is a working asset. The riskiest thing you can do to a working site is improve it — redesigns have broken more functioning lead generators than hackers ever have. Ugly-but-working is a maintenance case, not a project.
Gather this before deciding anything.
Five pieces of evidence, none requiring a consultant. Search your own company name — do you appear, and does the site show up for anything else? Open the site on a phone — most of your visitors already do. Find out when the platform was last updated, and by whom — if the answer is "unknown," that is the answer. Check who actually owns the domain, hosting and logins — a rebuild decision changes shape entirely if you first need a recovery. And ask the only commercial question that matters: has this site produced an enquiry in the last quarter? Those five answers usually make the verdict obvious — and when they conflict, that is genuinely what a professional review is for.
Whoever you ask, control for the incentive.
One structural caution: most advice on this question is free, and comes from companies that build websites — where every diagnosis has a way of concluding "rebuild." Our declared interest runs the other way, and we have priced it accordingly: the system review is a paid, written verdict across all four layers, and it is deliberately allowed to conclude that your site is fine as it stands — which a free quote-shaped diagnosis is not built to say. Whoever you engage, ask one question first: what would make you recommend against a rebuild? The quality of the answer is the quality of the advice.
Retrieved from https://amkatechnologies.com/insights/repair-rebuild-or-migrate
Published 30 July 2026. AMKA Technologies Sdn Bhd, SSM 202301041763 (1535682-T).