Site builder or custom build

What is the real difference, and why is this not just two options?

A site builder is a service that sells hosting, software and an editing tool together: you log in from a browser, take one of the ready made templates, drop in your text and pictures, and the site goes live on that company's servers. A custom build means the pages are written for your requirement and run on hosting you chose yourself.

Almost every comparison you read presents this as two points, and that is exactly where decisions go wrong. In practice there are four rungs and most businesses land on the second and third, not at either end of the range: a hosted builder, WordPress with a ready made theme, WordPress with a custom theme, and a program written from scratch.

Why does the distinction matter? Because cost, freedom and maintenance burden all three rise from one rung to the next, and someone who has only seen the two ends thinks they are stuck between cheap and limited on one side and expensive and free on the other. There is almost always a middle rung that does the job you actually want. If WordPress is unfamiliar, the previous lesson in this path explains the second and third rungs more closely.

Count the cost over three years, not over the first week

A cost comparison that only looks at day one always favours the builder, because nothing has been spent yet. The shape of the two costs is different, and it is the shape that has to be compared.

Builder cost starts low and never ends. It is monthly or yearly, and three things push it up: a higher plan for a capability you need later, every extra user, and add ons that each carry their own subscription. What the pricing table does not show is that in year one you pay for capabilities you do not have, and in year two you pay for the capability you have just started to need.

Custom build cost is the mirror image: one large amount at the start, then hosting and a domain which are small, and one line that is almost always left out, the maintenance budget. A site that gets built and then has nobody responsible for it is at its most expensive in year two, because its breakages get fixed at emergency rates.

We are not writing real market numbers here, because they move and putting them in a lesson manufactures durable misinformation. What we have measured is kept separately: the website cost guide. What you can do on this page is add up both columns for thirty six months and write the cost of leaving as its own line.

Ownership: when you want to leave, what comes with you?

This axis matters more than cost and gets asked about less. The question is simple: if you want to leave this service in two years, what exactly can you take? Four things are separate and each has its own answer.

• The domain. It must be registered in your account, not in the account of whoever built the site. This is the one item with no exceptions. If domains are unfamiliar, we explained them separately.
• The content. Most services offer an export and the quality of it varies enormously. One hands you a clean file; another hands you a file only it can read.
• The addresses. If the new site changes the URLs, everything those pages earned in search is lost unless a permanent 301 redirect is put in place. This is the one part of a migration that is not reversible.
• The design. That one never comes with you. A builder's design works on that system and nowhere else. Budget for rebuilding the look from scratch as part of any migration.

And one test that settles more than any argument: run the export now. Not when you have decided to leave. Create a trial account, add one page and one product, press the export button and open the file that comes out. If what comes out is not a file you can read yourself, you have your answer and no pricing table changes it.

Speed to a first version, and what it is actually worth

The builder wins here and there is no reason to dress it up. A few page brochure site is ready on a builder within days; the same site built custom takes weeks, because content, structure and design have to be decided before any code is written.

But the value of that speed is not what is usually sold. The value is not that you are online sooner; it is that you learn sooner. A site that goes up this month tells you by next month what people were looking for, which page they opened and where they made contact from. Then, if you decide on a custom build, you decide on data rather than on a guess. Commissioning a custom site before you know what people want from it is the most common way to throw money away in this field.

One warning belongs here: going up quickly is not the same as being temporary. Keep that fast version for two years and grow it, and it is no longer a prototype, while the cost of leaving it grows every month. If you know from the start that it is an experiment, put a date on it at the start, the date you will decide about it again.

The ceiling: where does a builder stop?

Every builder has a feature list, and outside that list nothing can be done. This is not a flaw in the product, it is its definition: in exchange for not having to maintain anything, you are limited to what they built.

The places that hit the ceiling earliest in Iran:

• Connecting to a local service: a payment gateway, a shipping system, accounting, or anything that exists only here.
• Control of the page head, where the canonical, structured data and language tags live. If the plan does not allow it, some SEO work is simply not available to you.
• Being multilingual at the URL level. Having the same page exist under a separate address per language is usually a paid add on or absent entirely.
• A data model that does not fit pages and products: bookings with rules, tiered pricing, or a listing with filters of your own.
• Control of speed. When you cannot decide which files load on a page, the main optimisation levers are not in your hands.

And now the counterpoint an agency usually does not say: most sites never reach that ceiling. A brochure site with a contact form and ten pages of content stays inside the feature list of any ordinary builder for its whole life. Building for a ceiling you will never touch is the most expensive mistake in this comparison, and it is repeated more than any other.

So when is it which?

We build and sell custom sites, so if we wrote here that a builder is always the wrong choice, we would be selling rather than teaching. The honest answer is this:

A builder is the right choice when you are testing an idea this quarter and do not yet know what the site has to do; when the budget is near zero; when what you want is a landing page or an event site; and when there is nobody to take maintenance on. Take that last one seriously: a custom site with nobody responsible for it is, six months later, worse than a plain builder.

A custom build is right when the site does the work rather than merely describing it; when it has to connect to a local service; when control of URLs, speed and the page head matters to you; and when somebody is there to keep it running. For WordPress specifically we answered the same question separately: where a WordPress build is right and where it is money thrown away.

A third option is real today and has a shape of its own: builders that generate the page with AI themselves. We looked at that group separately, because both its advantage and its troubles differ from an ordinary builder: AI website design, where it works and where it does not.

And if after all this you are still undecided, that hesitation is itself an answer. Hesitation means the requirements have not settled, and in that state the option that locks you in less is the better one.