At the bottom of the market there is no build fee at all: a hosted platform you set up yourself on a paid theme, a monthly subscription, your own evenings. In the middle, a freelancer configuring a bought theme for £800 to £2,500. A custom WooCommerce shop from me starts at £4,500, and most of the shops I am asked about land between that and roughly £7,000. Past £8,000 you are buying a team, and past £30,000, integration work I do not take on.
Two things before the detail. I am not VAT registered at present, so the figure I quote is the figure you pay, while an agency over the threshold adds twenty per cent. And the build fee is often the smaller half of the decision, because card fees, hosting, licences and somebody to change things run for as long as the shop trades. Every figure I charge is on the pricing page, and how the shop gets built is on the WordPress and WooCommerce page.
None of these six appears in a design mockup and most never appear in a quote either, which is how one shop ends up with two prices that look like they are for different jobs.
Four hundred products in a clean spreadsheet with one image each is an afternoon’s import. Sixty whose weights nobody recorded and whose photographs are on a phone is a fortnight, and most of that fortnight is not development. Ask what a quote assumed about who supplies the data.
Three sizes in four colours is twelve variants, each with a price, a weight, a stock count and possibly its own photograph. Sold by the square metre in boxes, with a wastage allowance and a minimum order, it is not a variant at all but arithmetic somebody has to write and test.
Weight bands by zone, a free-delivery threshold, collection in person, the one item the courier will not take, and the things that must never be charged postage at all: a gift voucher, a download, a dated ticket. Every one of those is a rule before it is code.
One standard rate on everything is straightforward. Mixed rates, zero-rated goods, alcohol duty or selling into the EU since Brexit are not, and none of it shows up in a mockup. Settle it in writing first, because retro-fitting a tax rule to a live catalogue is the expensive version.
Card fees on every order, a platform subscription if you are hosted, per-order app fees, and any booking or marketplace commission. The build is paid once. This is charged every month the shop trades, and on a busy shop it is the larger number by a distance.
Accounting, a till or stock system, courier labels, an ERP. Naming the actual software moves the price further than any design decision does. Two-way stock sync between a shop floor and a website is not a shop feature, it is software, and I price it as one.
Bands I see quoted, not survey data. There is no dataset behind them and I am not going to dress one up. My own figure is the only certain number in the table.
All three are builds I was the developer on during four years at a web studio in East Sussex, so they predate Kestrel. None is named and none is linked. The third gets a page of its own on vineyards and wineries.
WooCommerce, with the checkout path stripped back to the essentials. Weight, glass and an age-restricted product all constrain what the postage rules are allowed to be, and that gets settled before any design work starts.
A heavy catalogue kept genuinely quick to browse. Filtered browsing across that many options is a performance problem before it is a design one, and stone sold by area means the basket has to do sums a customer will trust.
One basket holding four incompatible things: a dated tour place with finite seats and nothing to post, a case that ships by weight, a recurring subscription, and a gift voucher that must never be charged delivery.
That third one is the argument in miniature: four products, one basket, and no two of them sharing a rule for stock, postage or renewal. It is why an online shop is not one price.
Take a shop turning over £15,000 a month with an average order of £60. That is 3,000 orders a year. Stripe publishes 1.5 per cent plus 20p for UK domestic cards at the time of writing, so the percentage costs £2,700 a year and the flat part £600. Call it £3,300 a year and £9,900 over three years, against a build at £4,500. The fees are more than twice the build, and they are charged whether the site is any good or not.
The flat 20p matters more than it looks. On a £60 order it is a third of a per cent. On a £6 order it is three and a third per cent, more than double the rate itself, so a shop selling small items should negotiate the flat fee rather than the headline percentage. Rates also vary by card type and by where the cardholder is, and every provider I have used charges more for a non-UK card. Take the figures off your own statement, not off my page.
If you have thirty clean products, one postage rule and nothing to integrate with, do not pay me £4,500. Set up a hosted platform, spend a fortnight on it, and put the money into photography and stock. A custom build earns its fee when the rules are complicated, not when the catalogue is large, and the crossover is the day the platform can no longer express a rule your business runs on. That argument is worked through on WordPress against Shopify.
My limits, plainly. I build shops on WooCommerce. I have built and shipped a Shopify app, so I know that platform from the inside, and I still do not sell Shopify theme builds. I do not do product photography, which on a shop is often the highest-value thing the money can buy. And a two-way stock sync with a till or an accounts system is a web app, from £7,500 after paid discovery.
If the catalogue is the problem rather than the checkout, a shop that has got slow is a Core Web Vitals job before it is a redesign, because a slow catalogue loses people before they reach the checkout, and that costs you sales whatever Google thinks of the site.
There is no single figure, and the honest range is wide. A hosted platform you set up yourself has no build fee at all, just a subscription. A freelancer configuring a bought theme is roughly £800 to £2,500. A custom WooCommerce shop from me starts at £4,500 and most land between that and about £7,000. A small agency with a designer, a developer and someone managing it is usually £8,000 to £20,000. Above £30,000 you are buying headless builds and ERP integration, which I do not take on.
Almost always because they assumed different things about the parts nobody discusses in a design meeting. Who cleans up the product data. How many variants there really are. Whether postage is one flat rate or weight bands by zone with exceptions. Whether the shop has to talk to your accounts or your till. Ask both quotes those four questions in writing. The cheaper one usually gets dearer once it answers them, and occasionally the expensive one turns out to have scoped work you do not need.
Less than the state of the data does. Four hundred products in a tidy spreadsheet with one image each is an afternoon's import. Sixty products with no recorded weights, descriptions living in a printed catalogue and photographs on somebody's phone is a fortnight of work, and most of that is not development at all. If you can supply a clean export with weights, dimensions, prices and images, you will pay less, whoever you hire.
More than most people expect, and they never stop. Stripe publishes 1.5 per cent plus 20p for UK domestic cards at the time of writing, and non-UK cards cost more with every provider I have used. A shop turning over £15,000 a month at £60 an order pays roughly £3,300 a year, so about £9,900 over three years, which is more than twice a £4,500 build. Rates move, so take the figures off your own statement rather than off my page.
It is cheaper to start and not necessarily cheaper to run. Shopify charges a monthly subscription and adds a further percentage on every order if you use a payment provider other than its own, while its own rate falls as you move up the plans. WooCommerce has no licence fee, but you pay for hosting and for somebody to build it. Compare the two over three years including fees, not on the build price. I build on WooCommerce and I do not sell Shopify theme builds.
Sometimes, and it is worth being careful about the word connect. A nightly one-way export of orders into accounting software is usually straightforward. Two-way stock sync between a till on a shop floor and a website, so that neither oversells, is not a shop feature: it is software, and I price it as a web app from £7,500 after paid discovery. If your system has no usable API, the honest answer may be that it cannot be done properly at all.
Hosting a real shop properly is not the £5 plan, so budget £20 to £80 a month depending on traffic. Plugin and add-on licences are commonly a few hundred pounds a year, and every renewal date is somewhere the shop can quietly break. Then card fees, which on a busy shop are the largest line by a distance. A care plan with me is £65 to £295 a month, or nothing at all and £65 an hour when something needs doing.
No, and on a shop it is often the highest-value thing the money can buy. Good photography of your own products will do more for sales than any front-end work I could do with the same budget. I will say that before quoting, and I will introduce you to a photographer without adding a margin. The same goes for copywriting beyond structure and editing. I am a developer, and the honest version of that is saying which parts are not mine.
Real numbers from your device and connection, not a screenshot of a score I chose. Nothing is sent anywhere. The measurement happens in your browser and stays there. If one of these ever reads amber on my own site, you will see it, which is rather the point. I audit these for £650.