Skip to content
Web Design

What a small business website should actually cost

Nobody will give you one number, and they are right not to. Here is what actually moves the price, the four kinds of project, and how to compare two quotes that look nothing alike.

by 4 min read
A close-up of a person working out a figure on a calculator, with technical drawings spread across the desk
Photo by RDNE Stock project on Pexels.

01 / The honest range

Why nobody will give you a straight number

Ask five agencies what a small business website costs and you will get five refusals to answer. That is not evasion for its own sake. “A website” covers a five-page brochure and a booking platform with payments, and quoting one price for both would be dishonest.

But you can be told what moves the number, and roughly where each kind of project sits. Here is how we think about it.

02 / The tiers

Four kinds of project, and what separates them

A presence site. Five to eight pages, your services, a contact form, editable by you. Nothing custom. Most trades, consultancies and clinics need exactly this and nothing more. Days of work, not weeks.

A marketing site. The same, plus pages built to convert: individual service pages, proof, a blog you will actually use, and search work so the pages can be found. This is where most of our projects land, and where the time goes from days into a couple of weeks.

A site that does something. Bookings, memberships, a catalogue, payments, a customer area. Now you are paying for logic, states and testing rather than pages. The cost is driven by how many things can go wrong, not by how many screens there are.

A product. Accounts, roles, billing, an admin panel. This is software with a marketing site attached, and it should be budgeted as software.

Most people asking the question want the second tier and have been quoted for the first or the third.

03 / What actually moves the price

Six things, and pages are not one of them

  • Whether the content exists. Writing your copy is real work. If we do it, it costs; if you do it, it costs you time. Either way somebody pays for it, and pretending otherwise is how projects stall.
  • How many templates, not how many pages. Forty pages on three templates is cheaper than eight pages that each look different.
  • Whether anything has to talk to anything else. A form that emails you is trivial. A form that creates a record in your CRM and books a slot is a different project.
  • Whether it takes money. Payments bring tax, refunds, failure states and legal pages with them.
  • How much design is decided. A brand with a palette and a logo starts from a foundation. Deciding the brand during the build is a second project running alongside the first.
  • Who has to approve it. Not a technical cost, but the most reliable predictor of how long anything takes.

04 / The number that matters more

What it costs to keep, not what it costs to build

The build is a one-off. The running cost is forever, and it is the part quotes go quiet about. Budget for hosting, a domain, any paid plugins or services, and someone to apply updates. It is not large, but it is not zero, and a site nobody maintains becomes a security problem within about eighteen months.

Then there is the cost that never appears on an invoice: every change you cannot make yourself. If updating a price means paying someone, that is a subscription with extra steps. Ask what you will be able to edit before you agree to anything.

05 / How to compare quotes

The cheapest quote is rarely the cheapest project

When two quotes are far apart, they are almost never for the same thing. Before you compare them, get every quote to answer the same five questions:

  • What am I able to edit myself afterwards?
  • Who owns the files, the domain and the hosting account?
  • What happens to my existing URLs and rankings?
  • What is included after launch, and for how long?
  • What is explicitly not included?

That last one separates a considered quote from an optimistic one. Anyone who has done this before knows where projects overrun and will tell you.

06 / Honestly

When you should spend less than you were planning to

If you are pre-revenue, testing whether anyone wants the thing, or your customers all come from one referral partner, buy the cheapest credible site that answers the obvious questions, and spend the difference on finding out whether the business works. You can rebuild later from a position of knowing something. We have told people this and lost the work, and it was still the right advice.

Tell us what you are trying to build and you will get a range with the assumptions written next to it, and a scope sketch either way.