What a small business website should actually cost

A website is not one product. It is a bundle of decisions about design, content, functionality, integration and who does the work, and each of those decisions moves the number. Until they are settled, a price is a guess dressed up as a quote.

So rather than give you one figure, here is how we band it, and then the specific things that push a project up or down inside those bands. If you know which of them apply to you, you can read any quote you are given and tell whether it is realistic.

These are our published ranges, not a teaser figure. We agree an exact price in writing after a short scoping call, and that price is what you pay.

You can get a website for a few hundred pounds. What you buy at that price is a template filled in quickly, usually on a platform that charges you monthly forever, with your content written by you and nobody thinking about search, speed or how people get from the page to contacting you. For some businesses that is the right call. A one person operation testing an idea does not need an eight thousand pound site.

The honest problem with the cheap end is not the design. The parts which decide whether a site earns anything, the structure, the speed, the words, the route to an enquiry, are exactly the parts that get skipped at that budget. You end up with something that exists rather than something that works.

Above the bands you are paying for scale rather than polish. Thousands of products, several languages, or a business where the website is the product rather than a shop window for it. At that point the number stops being about pages and starts being about complexity.

When a project runs late or costs more than the first conversation suggested, the cause is content far more often than code. Design and build can be planned. Waiting for someone to write the words for the fourteen service pages cannot.

There are three states to be in and they price very differently. Finished copy, real photographs of your own work, logo files and product data in a spreadsheet puts you at the bottom of the band. Rough notes and stock photographs puts you in the middle. Needing someone to interview you, write everything, arrange photography and build the product data puts you at the top, reasonably, because that is weeks of skilled work.

This is the one thing you can genuinely change to lower a quote. Getting your content ready before you commission a build tends to save more money than any amount of haggling over the design, and it removes the most common reason projects stall for months.

The most expensive word in a web project is just. It can just take a payment. It can just push the enquiry into our CRM. It can just show live stock. Each of those is a whole piece of work, and the cost depends almost entirely on the system at the other end rather than on the website.

A modern system with a documented interface is a known quantity. An old piece of industry software with no interface, or a supplier who will not give you access to your own data, is not, and the fair way to price it is to look first. Anyone who quotes an integration confidently without seeing both ends is guessing.

Decide too which integrations you genuinely need at launch. A great many can wait until the site is live and earning, and staging them that way tends to get you online sooner for a smaller first cheque.

The difference is not the number of pages. It is that a shop has to be right rather than merely convincing. A brochure page that says something slightly wrong is embarrassing. A shop that calculates VAT wrongly, oversells stock it does not have, or drops the last step of the checkout, costs you money and customers every single day it is live.

So a shop carries work that a brochure site simply does not: payment providers and the testing that goes with them, shipping rules including the awkward ones about weight, size and where you will and will not deliver, VAT treatment including sales to other countries, stock that has to stay accurate, returns and partial refunds, and product pages structured so they can be found in search rather than just existing.

The product data itself is the sleeper cost. A hundred products with clean codes, real descriptions and decent photographs is a straightforward import. A hundred products spread across a spreadsheet, a supplier catalogue and someone's memory, with three codes for the same item, is a data project before it is a web project. Be honest with yourself about which of those you have.

A build price is a one off. Running a website is not, and a quote that hides the ongoing cost is not cheaper, it is just less clear.

When you have a cheap quote and a dear one, do not start by asking why the dear one is dear. Start by asking both of them the same short list of questions, because the answers usually explain the gap entirely.

Who writes the words. Who owns the code, the domain and the hosting when it is finished. Can you edit the pages yourself, and can you see that being done before you sign. What happens to the search rankings on your current site. What is the monthly cost after launch, in total, including everything. What is explicitly not included. And what happens if you want to leave in two years.

In our experience the gap between two honest quotes turns out to be scope rather than greed, and once both are answering the same questions they move a lot closer together. If you want a straight read on a quote you have been given, send it to us. We will say what we think it covers and what it does not, including when it looks like a fair price.