A website should be rebuilt when what it is built on can no longer do what the business needs, and not merely when it has started to look old. The distinction matters because a rebuild is the most expensive and riskiest option available, and most of the reasons people give for it are solved more cheaply by something smaller. Four reasons genuinely justify starting again, and one popular reason does not.
One: the business is no longer the one the site describes
A new offer, a new kind of buyer, a service dropped or split in two. When a company's real offer has moved, the site's structure — its pages, its navigation, the order in which it tells the story — describes a company that no longer exists. Rewriting the text helps for a while, but if the pages themselves are the wrong pages, you end up forcing the new business into the old skeleton.
Two: nobody can change it
If a headline edit needs a developer, if a change in one place breaks another, or if the only person who understood the site has left, it has stopped being a working asset. The test is practical: ask for a small change and time it.
A site where small changes are slow gets edited less, and a site that is edited less goes stale — slowly at first, then visibly. At that point the cost of the old site is not what you pay to keep it running but what it stops you from saying.
Three: it fails in ways patches cannot fix
Speed, accessibility and search structure are mostly decided by foundations. A site built on a platform that is no longer maintained, or on a layout system that nests every element many levels deep, can be tuned but not rescued. When each fix means fighting the structure, the sum of the fixes is usually more than a rebuild would cost, and the result is still worse.
Four: it was never built for where you are going
A second language, a shop, a large content archive, a partner area. Some of these can be added to an existing site; others change its structure so deeply that bolting them onto a site that did not expect them produces something that works and that nobody likes to touch. Structure is far cheaper to decide first than to retrofit, and the honest moment to ask the question is before the new requirement arrives, not after.
The reason that is not on the list: it looks dated
A site that looks old but loads quickly, reads clearly, ranks, converts and can be edited has a visual problem, and a visual problem has a visual solution: restyle the surface and keep what is underneath. That costs a fraction of a rebuild and does not put the site's existing search standing at risk.
Rebuilds are often sold on appearance because appearance is what a presentation can show and foundations are not. Be most suspicious of a rebuild whose entire case rests on how the new design looks.
Restyle, repair, rebuild
These are three options in ascending cost and risk. A restyle changes how the site looks. A repair fixes specific defects underneath. A rebuild replaces the foundation. Ask any studio that proposes the third which of the first two it considered and why it rejected them. If it cannot tell you, the proposal was written before the diagnosis.
What a sound rebuild proposal contains
If the diagnosis says rebuild, the proposal should show its working. It lists the existing addresses and says which survive, which move and which are retired, because that decides whether the site keeps its standing in search. It says what content is carried over and what is rewritten, who owns each, and how the old and new sites will be measured against each other.
It also says what the rebuild will not do. A rebuild that promises to fix traffic, conversion, speed and brand at once has not been scoped. A good one names the two or three problems it exists to solve and states how you will know whether it solved them.