A patched together website can run for years. It just runs badly.
Most businesses do not decide to rebuild a website on purpose. They decide after the tenth small fix stops holding, once patch after patch has quietly cost more than a rebuild would have cost outright.
I look at old sites for a living. The pattern repeats enough to give it a name.
The Patch That Doesn't Hold
A plugin gets added to fix a slow form. Another gets added to compress images. A third handles a popup nobody remembers approving.
Each fix solves its own problem. None of them talk to each other.
The site slows down.
Something breaks every few months, and the fix is another plugin stacked on the ones already there. This is not a redesign problem. It is a foundation problem, and a foundation problem does not go away with new photography.
What Counts as a Real Signal
Some signs matter more than others. A slow site matters. A site running the same code as eight years ago matters more.
Here is what I actually check on an old site.
- The text cannot be updated without calling a developer.
- Pages take more than four seconds to load on a phone.
- The design has not changed since a different logo was in use.
- Every new feature needs a new plugin that slows the whole site further.
One of these is worth a fix. All four together are not a website with problems. They are a website that was never built with a plan, patched one decision at a time.
What a Rebuild Actually Buys
A rebuild is not a new coat of paint. It is a new foundation the business can build on instead of around.
The gap shows up first in speed. It shows up next in what a visitor can actually do on the page, and whether that page turns a visitor into a lead instead of losing them before the scroll.
Then it shows up in the code underneath. Real code reads clean to a browser and to the software that renders the page. Five years of stacked plugins does not, and that costs a business twice, once with the person on a phone and once with the software reading the page for Google and for AI tools like ChatGPT.
Our post on cinematic web design versus template websites covers the design side of that gap. This is the part underneath it.
What I Wouldn't Touch Yet
Not every old site needs to start from scratch. A site that loads fine, works fine, and just looks dated has a design problem, not a foundation one. New photography and a cleaner layout can close that gap for a fraction of a rebuild.
What I wouldn't waste money on is a full rebuild for a site whose only real complaint is that the owner is tired of looking at it. That is a redesign. Price it and scope it like one.
The mistake runs in both directions. Some businesses rebuild too early, chasing a trend. Others patch too long, treating a foundation problem like a design one.
What the Old Code Costs a Crawler
A stack of plugins does not just slow a page down for a person. It buries the actual content under code a crawler has to work harder to parse.
Structured data added by one plugin often conflicts with structured data added by another. Google and AI tools like ChatGPT read whichever version loads, and it is not always the correct one.
None of this shows up in a browser. The site looks fine on screen. It just does not read the way the business assumes it does to something trying to understand what the page is actually saying.
The Right Time to Decide
A rebuild is worth it when the foundation, not the decoration, is the problem. Slow load times no plugin fixes. Content that cannot be updated without calling someone. Structured data that was never built correctly the first time.
Fix the foundation, and the surface gets easier to keep current after.
If three or more of the signals above are true, stop paying for patches. A rebuild that redirects every old page correctly costs less over two years than another year of small fixes on a site that was never built to hold them.
UM Media builds custom websites from the ground up, inside the Web Experiences process, scoped to what the site actually has to do.