What Actually Drives the Cost of a Business Website

Ask three developers to quote for "a company website" and you'll get three very different numbers. That's not usually because two of them are trying to overcharge you. It's because "a company website" describes almost nothing, and each of them has quietly filled the gaps with different assumptions.

Here's what actually moves the number, so you can have a more useful conversation.

Driver one: how many distinct things it does

Pages are cheap. Capabilities are not. Ten pages of text cost far less than one page where a customer picks a date, checks availability, pays, and gets a confirmation.

A useful way to think about it: count the things a visitor can do, not the things they can read. Browsing is one capability. Searching and filtering is another. Booking is another. Paying is another. Logging in to see their own history is another — and that one is bigger than it looks, because accounts bring password resets, permissions, and a duty of care over personal data.

Most quotes that come in surprisingly high are responding to a capability the business mentioned in passing and didn't think of as a feature.

Driver two: integrations

Every system your website has to talk to is a moving part you don't control. Payments, accounting software, an existing stock system, a delivery partner, an SMS provider, a CRM.

Integrations are expensive for reasons that aren't obvious from outside:

The other system's documentation is often incomplete, and finding out how it really behaves takes time.
You have to handle it being down, slow, or returning something unexpected — without your site falling over.
Access and credentials frequently involve a third party's approval process on their timeline, not yours.
When it changes, you have to change too, forever.

One integration is a feature. Five integrations is the project.

Driver three: whether your content is ready

This is the most underestimated line item, and it's the one you have the most control over.

If you arrive with final text, real photographs, your logo in a usable format, and a clear idea of what each page should say, the build moves quickly. If the site is designed around placeholder text that later gets replaced by content twice as long, layouts have to be reworked — and rework is paid for twice.

The single cheapest thing you can do to reduce the cost of a website is to have the content finished before the build starts.

It's also the most common reason projects run late. Development finishes; the site waits three months for photographs.

Driver four: how bespoke the design is

There's a real spectrum here, and all of it is legitimate:

Template-led. A proven layout adapted to your brand colours, typography and content. Fast, predictable, and perfectly respectable for many businesses.
Designed around your content. Layouts built for what you specifically need to communicate, rather than fitting your business into someone else's structure.
Fully bespoke with custom illustration, photography or motion. Appropriate when the site itself is a primary sales asset and needs to be distinctive.

The mistake isn't picking the cheaper end. It's paying for the expensive end while supplying content and photography that undercut it.

Driver five: what happens after launch

A website is not a thing you buy once. Someone has to renew the domain, keep the hosting running, apply security updates, take backups, and fix what breaks. If that isn't in the quote, it isn't free — it's just unassigned, and it usually lands on whoever in your office is considered "good with computers".

Ask explicitly:

Who holds the domain registration, and in whose name? It should be yours.
What's included after launch, and for how long?
What happens if I want to change something myself?
If we part ways, what do I walk away with?

That last question is worth asking early. A website you can't take with you isn't really yours.

The thing that saves the most money

Not the technology choice. Not haggling on the day rate. It's being able to edit your own content.

If every price change, new team member, blog post or product update requires a developer, you'll pay for that steadily for years — and worse, you'll stop making changes because it's a hassle, and the site will slowly go stale. A proper admin area costs something up front and stops that happening.

A note on the cheapest quote

The lowest number is sometimes the right choice, and we'd never argue otherwise on principle. But it's worth understanding what usually produces it, because there are only a few possibilities: a smaller scope than you described, a template you'll recognise on other sites, less time on the parts you can't see, or an assumption that the shortfall gets recovered later through change requests.

Any of those can be perfectly acceptable if you've chosen it knowingly. The problem is choosing it unknowingly, then discovering the gap at the point where switching is most expensive — halfway through, with content written and a launch date announced.

So when quotes differ a lot, the useful question isn't "why is that one so expensive?" It's "what is the cheaper one assuming that the dearer one isn't?" That question usually gets a straight answer, and the answer tells you which one you actually want.

How to ask for a quote that means something

Instead of "how much for a website", send this:

What the business does, and who the site is for.
The three things you most want a visitor to be able to do.
Any system it must connect to.
What you'll be able to supply — text, photos, logo — and when.
What you want to be able to change yourself afterwards.
Whether there's a date that actually matters, and why.

You'll get comparable quotes instead of three numbers measuring three different projects. And if someone quotes a flat price off a one-line brief without asking any of this, that's worth noticing too — either they've made assumptions you haven't seen, or the difference will surface later as a change request.

If you'd like a straight breakdown for what you have in mind, tell us what you're trying to do. We'd rather scope it honestly than quote a number that changes later.