Joomla upgrade or rebuild: how to decide

A Joomla 3 site has been out of support since August 2023, and a Joomla 4 site since October 2025. Sooner or later the hosting company retires the PHP version the site depends on and the question can no longer wait: upgrade the site you have, or build a new one and move the content into it.

Both routes end at a current Joomla version. They differ in what they keep, what they cost and what can go wrong on the way.

What the two routes are

An upgrade keeps the existing site and its database and moves it forward with Joomla's own update component. It goes in steps, because Joomla does not skip major versions: 3.10 to 4.4, then 4.4 to 5.4, then on to 6. Articles, categories, menus, users, passwords and addresses stay where they are. Everything around them, the template and every extension, has to be compatible at each step or be removed before it.

A rebuild starts with a fresh installation of the current version. The design is built as a new template, the needed extensions are installed new, and the content is imported from the old site. Nothing old comes along unless it is carried across on purpose.

Question 1: what happens to the extensions

List every installed extension and find out, for each one, whether a version exists for the Joomla version you are heading to. Joomla's update screen runs a pre-update check that does part of this, but it only knows what the extension's developer has declared, so the list needs checking by hand.

Extensions fall into three groups. Those with a current version are updated before or during the upgrade. Those that are abandoned but replaceable get a successor, and their data has to be moved into it. Those that are abandoned and hold important data, an old directory component or a booking system, are the expensive ones. Three or four of those and the upgrade route loses most of its advantage, because the hard work is the same on both routes.

Question 2: what happens to the template

A Joomla 3 template does not run on Joomla 4 or later. The underlying front-end libraries and the way templates are structured both changed. If the template's vendor offers a version for the new Joomla and the site moves onto it with only minor settings and styling fixes, we keep it. If not, there are two choices: a new template, or the existing one reworked by us until it is compatible with the new version. The second keeps the look the visitors know. The first is the better choice when a redesign was due anyway.

This matters for the decision more than anything else. Template work is often the largest single cost, and on most Joomla 3 sites it is needed on both routes. Once that is accepted, the remaining question is narrower: is the existing database worth keeping?

Question 3: how much content, and how tidy

A site with thousands of articles in a sensible category structure, with custom fields, tags and years of internal links, is a strong case for upgrading. Everything keeps its ID and its address, and nothing has to be mapped.

A site with 25 pages is a weak case. Copying 25 pages into a fresh installation takes less time than preparing, rehearsing and checking a three-step upgrade.

Question 4: how old is the database

Many Joomla 3 sites began life on Joomla 1.5 or 2.5 and were migrated once already. They carry tables from extensions removed long ago, inconsistent permission records and settings nobody remembers choosing. An upgrade brings all of it along unless someone removes it, and some of it surfaces as errors halfway through the process. Cleaning the database of that leftover material is part of our upgrade work for this reason.

A database that started on Joomla 3 and has been looked after needs little of this. One that has been through two major migrations and a change of developer deserves a careful look before anyone commits to upgrading it.

Question 5: is there custom code

Components, modules or plugins written specifically for the site have to be ported to the new Joomla version on either route. We examine the code and quote the port separately, and the cost is the same on both routes. What differs is the risk: on the upgrade route, unported custom code can block the update itself, so it has to be finished first. On a rebuild it can be ported in parallel.

Question 6: users and addresses

An upgrade keeps user accounts and passwords without any extra work. A rebuild can keep them too, but only if the import is written to carry them across.

Addresses behave the same way. An upgraded site keeps its URLs as long as the routing settings are left as they were. On a rebuild, the old addresses have to be recreated or redirected one by one. For a site that earns its visitors through search, a full list of existing addresses is the first thing to produce on either route.

How the answers add up

There is no formula. We decide it site by site, after looking at the answers above. Some patterns hold often enough to be useful.

Upgrading in place tends to win when the site has a lot of structured content, registered users, extensions that are still maintained and a database with a clean history.

Rebuilding tends to win when the site is small, when a redesign is planned anyway, when the key extensions are dead, or when the database has been through a lot. It is also the moment when it is reasonable to ask whether Joomla is still the right platform for the people who run the site. We work with both Joomla and WordPress and treat that as an open question with no preferred answer.

There is a middle route that is often the best one for medium-sized sites: a fresh installation, a new template, and a scripted import that keeps the article IDs, aliases, categories and users from the old database. It gives the clean base of a rebuild and keeps most of what an upgrade would keep.

The cost of waiting

An unsupported Joomla version keeps working until something underneath it changes. Usually that is the PHP version, and the first sign is a 500 error after a PHP update. A decision made that week is made under pressure, with the site down. The same decision made six months earlier is made on a staging copy with nobody watching.

Whichever route fits, it should be rehearsed on a copy of the site before anything happens to the live one. Our upgrade and migration service starts with exactly this assessment: an inventory of the site and its dependencies, and a written plan with an estimate for the route that suits it.