Website cost is the one number every proposal gives you, and the least useful number in the document.
Say you have three quotes in front of you. Six thousand pounds, eighteen thousand, forty thousand. All three describe roughly the same thing in slightly different language. Not one of them tells you what the site will earn.
So the decision gets made on gut feel, or on whichever designer you got on with, or on whichever number hurts least this quarter.
There is a better way. It takes twenty minutes with numbers you already have, and it gives you a website cost you can defend to your bank.
TL;DR: Do not start with what a website costs. Work out the extra gross profit the site has to generate, then spend no more than about two years of that. Your margin sets the ceiling, not your revenue, and most of the difference between a cheap quote and an expensive one has nothing to do with design.
Website cost is the wrong place to start
You would not buy a machine by asking what machines cost. You would work out what it needs to produce, then see which machines clear that bar at a price you can justify.
A website is the same kind of purchase and almost nobody treats it that way. It gets bought like a brochure, on look and feel, then judged like an investment two years later when somebody asks what it ever did.
The question worth asking is not what a website costs. It is what this website has to earn to be worth building.
Work out what the site has to earn
Four numbers, all already in your business.
- Enquiries the site produces a month.
- How many of those you convert into paying customers.
- Your average order value.
- Your gross margin on that work.
Take a business getting 20 enquiries a month from the site, closing one in five, at an average order of £5,000. That site sits behind four orders a month. £20,000 of revenue a month, £240,000 a year.
Now assume a rebuild lifts enquiries by 20%. Four more enquiries a month, so 0.8 of an extra order, so roughly £4,000 a month. Call it £48,000 of extra revenue a year.
That is the number every proposal wants you to look at, and it is the wrong one.
Your margin sets the ceiling, not your revenue
You cannot spend revenue. You can only spend what is left of it.
On a 20% gross margin, that £48,000 of extra revenue is £9,600 of actual profit. Two years of it is about £19,000. So £20,000 is a defensible website cost and £40,000 is not, however good the pitch was.
Run the same business at a 40% margin and the same lift is worth £19,200 a year. Now £38,000 is defensible and the expensive quote is back in play.
Two businesses, identical traffic, identical quotes, and the right answer differs by twenty thousand pounds. No agency can tell you what your website should cost without asking about your margin. Be wary of any that quotes before it does. Our piece on what a business your size should actually spend applies the same test to ongoing marketing.
Your accountant already treats it as an asset
Worth knowing, because it changes how the spend should feel.
HMRC’s own guidance treats the original cost of building a website as capital expenditure, and the ongoing cost of updating it as a revenue expense. It uses a shop window to explain the split: constructing the window is capital, changing the display from time to time is revenue (HMRC Business Income Manual).
So the tax authority thinks you are buying an asset with an enduring benefit. Most owners think they are buying marketing. That gap explains a lot of bad website decisions, in both directions: underspending on something that has to last four years, and overspending on a display that only needed changing.
Your accountant will tell you how it applies to your accounts. The point is that a website is closer to plant than to an advert.
What actually makes one quote three times another
Owners assume the website cost gap is design quality. It very rarely is. Three things drive most of it.
- Page types, not page count. Forty pages built from six templates is a small job. Twelve pages that all look different is a big one. Ask how many distinct templates are in the quote.
- Connections to systems you already run. Pulling live stock, pricing, quoting or CRM data into the site is where budgets go. If it is in the quote, it should be itemised. If it is not, ask what happens when you want it.
- Who writes the words. The cheap quote almost always assumes you will. Most owners never do, the project stalls for four months, and the site launches with the old copy on it.
Get those three answered and the quotes usually stop being comparable in the way you thought.
What good looks like in numbers, and what it does not buy you
There are published thresholds you can hold a supplier to. Google’s Core Web Vitals set the bar at 2.5 seconds or less for the main content to appear, 200 milliseconds or less for the page to respond when someone taps something, and a layout stability score of 0.1 or less, measured across the slowest quarter of real visits.
Put those three numbers in the contract. They are objective, measurable after launch, and cost nothing to ask for.
Now the honest part. Hitting them will not lift your rankings on its own. Google says plainly that there is no single page experience signal, and that it seeks to show the most relevant content even where the page experience is sub-par (Google Search Central).
So treat a green score as a hygiene standard, not a growth strategy. Paying a premium to chase it is one of the easier ways to inflate your website cost for nothing.
The mobile assumption that quietly costs money
You will be told the site must be designed for mobile first, as though that settles it. Check before you accept it.
Across all UK web traffic in August 2026, desktop was 51.6% and mobile 48.4%, roughly an even split (StatCounter). For plenty of B2B firms, where people visit from a desk during working hours, the desktop share is far higher than that.
Your own analytics will tell you your split in a minute. If 70% of your enquiries come from desktop, a site designed as a phone experience and stretched to fit a monitor is costing you money.
When the honest answer is do not buy one
Sometimes a rebuild is the wrong purchase, and you will not often hear that from someone selling rebuilds.
If enquiries are healthy but poor quality, that is usually a copy and qualification problem on five or six pages, not a structural one across forty. If your site simply looks tired, ask what a new one would change commercially before you spend.
A couple of thousand pounds spent fixing your five most visited pages, the enquiry form and how quickly you respond will often tell you more than a full rebuild would, and it tells you sooner. We have written about why website forms go unfilled when the traffic itself is fine.
If that work lifts your enquiries, you have proved the mechanism and earned the right to spend properly. If it does nothing, you have just saved yourself the cost of finding out the expensive way.
How much should a website cost for a small business?
There is no standard price, and any figure quoted before someone has asked about your margin is a guess. The useful answer is a ceiling you can calculate yourself. Work out the extra gross profit the site should generate in a year, then spend no more than about two years of that. If the site should add £10,000 of profit annually, £20,000 is defensible and £45,000 is not.
How do I work out whether a new website will pay for itself?
Take your current monthly enquiries, the share you convert, your average order value and your gross margin. Estimate a realistic percentage lift in enquiries, apply it through those four numbers, and you have the extra annual profit. Divide the quoted website cost by that figure and you have your payback in years. Anything beyond two to three years deserves a hard second look, because most sites get replaced inside four.
How long should a website last before it needs replacing?
Plan on four to five years for the build itself, with content and pages updated continuously throughout. Looking dated is rarely a good enough reason to replace a site. Genuine reasons are that it can no longer do a commercial job you now need, that it cannot be updated without a developer, or that it is measurably slow or broken on the devices your customers actually use.
Is a new website a capital cost or a marketing expense?
HMRC’s guidance treats the original cost of creating a website as capital expenditure, and regular updates to it as revenue expenses, using the analogy of building a shop window versus changing the display. How that applies to your accounts is a question for your accountant. Commercially, the useful takeaway is that a website behaves more like a piece of equipment with a working life than like an advert you run once.
