Build your own website, or hire someone?
Four routes, and the three questions that decide between them.
Every business owner who needs a website arrives at the same fork, usually late in the evening: do I make this myself, or do I pay someone? The honest answer depends on three things, and none of them is your budget.
There are four routes open to you. Each one is genuinely the right answer for someone. This is how to work out which someone you are.
Route one: build it yourself in a website builder
A drag-and-drop builder is the cheapest route in money and the most expensive in attention. You will spend somewhere between one and four evenings on your first version, and the result will look the way you arranged it.
What people underestimate is the second week. The site is live, a price changes, and you have to remember how the editor worked. Most abandoned websites were not badly built — they were built by someone who never came back.
Choose this if you enjoy the making, or if the site is genuinely temporary: one event, one campaign, one season.
Route two: describe it and let AI build it
You write a description of your business, and software produces a working site from it. Your part takes about an hour of your time rather than several evenings, and you never open an editor.
The trade is control at the margins. You are not placing elements on a page; you are saying what the page is for. For a five-page site that has to exist, look professional and be findable, that trade is usually a bargain. For a site whose exact layout is the product — a portfolio where the arrangement is the work — it is not.
This is the route ChatBuild takes, so read the next two sections with that in mind: there are jobs on this list we are the wrong answer for.
Route three: hire a freelancer
A freelancer gives you a person. That is the whole value, and it is a real one: someone who has seen your kind of business before, who will tell you your opening paragraph is unclear, and who can be phoned.
The risk is also that it is one person. Holidays happen, priorities shift, and a small change six months later competes with their current project. Ask two questions before you sign anything: what does a change cost after launch, and who owns the domain? If the domain is registered to them rather than to you, walk away.
Route four: hire an agency
An agency is the right answer when the website is not really the deliverable. If you need research, a brand, a photographer, integrations with systems you already run, and someone accountable for the result, that coordination is what you are buying, and it is worth paying for.
You are also buying a process. Expect a discovery phase, a design round and a build, expect it to take weeks rather than days, and expect a one-off fee in the thousands plus a maintenance contract afterwards. For a site that will carry real revenue, that can be the cheapest option on this page. For a one-person business that needs to be findable by Thursday, it is not.
The three questions that actually decide it
Skip the comparison tables. Answer these instead.
How often will this change? A site you touch twice a year and a site you touch twice a month are different products. Frequent change is where do-it-yourself and one-off builds both quietly fail — the first because you forget how, the second because every change is an invoice.
Is the layout the product? If a visitor's experience of the arrangement is what you sell, you need design control and therefore a designer. If the site's job is to explain what you do and let someone contact you, layout control is a cost pretending to be a feature.
Who will own it in two years? Not who builds it — who holds the domain, who can move it, who can get the content out. This is the question nobody asks and everybody regrets.
One thing to check whatever you choose. Your domain should be registered in your own name, not your builder's. It is the one part of a website that is genuinely yours, and the one part that is painful to get back later. How ChatBuild handles domains.
Where each route goes wrong
Every route on this page has a failure mode, and knowing it in advance is worth more than a feature comparison.
- Builder: the site goes stale because editing it is a task, and tasks lose.
- AI: you accept the first version on subjects you should have argued with — particularly the sentence about why someone should choose you.
- Freelancer: the relationship, not the site, becomes the dependency.
- Agency: you get an excellent site and a change process too slow to use.
Notice that three of the four are about what happens after launch. Launch is the part everyone plans for and the part that matters least.
Compare three years, not the quote
The quote is the least useful number in this decision, because it describes the cheapest moment of owning a website. Work out three years instead, for each route you are seriously considering.
Four lines, and you can fill them in from any proposal:
- What you pay to get it live. A build fee, a subscription with no build fee, or nothing but your own evenings.
- What you pay every month whether or not anything happens. Hosting, the domain, a maintenance contract, a subscription.
- What a change costs. The number that decides whether your site is still accurate in year two. Ask for it as money and as days.
- What leaving costs. Can you export the content, and is the domain in your name? A route that is cheap to enter and expensive to leave is not cheap.
ChatBuild is a subscription with no build fee, and Web Starter is €79 per month excl. VAT with changes included up to the tier's limit — so for us, lines one and three are the interesting ones and line two is the whole subscription price (a domain you buy through us is its own yearly payment). An agency quote typically inverts that: a one-off build in the thousands, plus a maintenance contract, plus a rate for changes. Neither shape is better in the abstract. They are better at different change frequencies, which is why the first of the three questions above is the one that matters.
What to do next
Write down, in one paragraph, what your site has to achieve in its first year — be found, explain, take bookings, sell. Then put a number next to how many times you expect to change it.
If that number is low and the paragraph is simple, almost any route works and you should pick the cheapest one you will actually finish. If the number is high, choose on change cost, not on build cost. And if the paragraph turned out to be complicated, that is useful information too: it means the website is a project, and projects want a person.
We wrote the honest version of what our own route does well and where it leaves people stuck in the limits of AI website builders. If you would rather see the sequence from idea to live first, that is here.
Weighing up a web design agency? Compare the costs over one and three years, the timelines, what you keep when you leave and when an agency is the better choice: web design agency vs website subscription.
Frequently asked
- Is a website builder cheaper than a subscription?
- In money, usually yes; in attention, usually not. A builder costs you evenings to set up and a recurring decision to maintain. The honest comparison is three years of both, including the changes you will want to make.
- How do I know if my project is too complicated for AI to build?
- Try to brief it on one page. If the description keeps sprouting exceptions and rules, those exceptions are the project, and they need a person who owns them rather than a tool that generates them.
Sjors Hooijen Builds ChatBuild and writes about getting a business online without turning it into a project. Questions about this article? Email hi@chatbuild.com.