Ecommerce Website Cost: What You Pay For and What Moves the Number

An ecommerce website cost is really three numbers pretending to be one: what you pay to build it, what you pay every month to keep it running, and what every order costs you in platform and payment fees. Quotes usually show the first and stay quiet about the other two, which is why a €4,000 shop and a €40,000 shop can both be described as "an online store".

This breaks the number apart. Product count, checkout complexity, integrations and content are what actually move the build price, and none of them appear in a page count. After that come the fees, which are invisible on day one and material by month twelve.

The bands below are what we quote and what we see quoted across the European market. Where a figure comes from a platform or a research body rather than from us, it is named and linked.

Four things move the number, and page count is not among them. How many products you carry and how they vary, how complicated the checkout has to be, how many systems the shop has to talk to, and who writes the product content. A 40-product shop with one price and one shipping rate is cheaper to build than a 12-product shop where each item has sizes, materials, engraving and three tax treatments.

Product variation is the one that surprises people most. A single product with five sizes and four colours is twenty saleable items, each needing stock, images and a price, and the admin screens have to make that manageable for whoever runs the shop on a Tuesday afternoon.

Integrations are the second surprise. Every connection to an accounting system, a warehouse, a supplier feed or a shipping carrier is a small project of its own, with its own failure modes and its own testing.

Four honest bands cover almost every store we are asked to quote. They assume product photography and product copy are either supplied by you or written into the project as a separate line, because that is the largest single variable and hiding it is how quotes go wrong.

Read the bands as ranges, not as menus. A shop lands in a band because of variant count, checkout rules and integrations, and the same visual quality is achievable in all four.

Platform choice changes where the money goes more than how much of it there is. A hosted platform moves cost from the build into a monthly subscription and a per-order percentage. An open-source stack moves it the other way, more up front and less per order, with the hosting and updates now your responsibility.

The break-even is a matter of order volume, not opinion. Below a few hundred orders a month, hosted is almost always cheaper in total. Above a few thousand, the percentage starts to fund a lot of development.

The second question is who edits the shop after launch. A merchandiser who can add a product, change a price and run a promotion without a developer saves more money over three years than most platform decisions do.

Per-order fees are the part of ecommerce website pricing that never appears in a build quote and quietly becomes the largest line. They come in two layers: what the platform takes and what the payment provider takes, and both are percentages, so both scale exactly with your success.

Shopify's published pricing puts its plans at $39, $105 and $399 a month with online card rates of 2.9 %, 2.7 % and 2.5 % plus $0.30 per transaction, and adds a further percentage if you use an outside payment gateway. In Europe, Stripe's standard rate for European cards is 1.4 % plus €0.25, with non-European cards priced higher.

Run those against your own numbers before choosing. At 500 orders a month averaging €60, a one-point difference in total fee rate is €3,600 a year, which is most of a small build.

The arithmetic that matters is total cost per order, not the headline percentage. Add the platform's cut, the payment provider's percentage, the fixed fee per transaction and any gateway surcharge, then divide by your average order value. Shops with a €25 basket feel fixed fees far more than shops with a €250 basket, and that single fact should shape the platform decision more than any feature comparison table.

Quotes and invoices part company in four predictable places, and none of them are anybody acting in bad faith. Product content gets assumed. Photography gets postponed. The tax and shipping rules turn out to be more specific than the first conversation suggested. And someone discovers in week six that the accounting system needs a field nobody mentioned.

The fix is to put each of those in the quote as its own line with its own price, even when the answer is "client supplies". A line that says zero is still a line, and it stops the conversation in month three from being a surprise.

Migration is the fifth. Moving an existing shop is not a discount on a new build, it is a new build plus a mapping exercise: old URLs to new ones, order history, customer accounts, and the redirects that keep the search rankings you already paid for.

A shop is not a website with a checkout bolted on; it is a system that other systems depend on, and it needs a budget after launch. As a rule of thumb, plan 15 % to 25 % of the build cost per year for hosting, platform fees, security updates, plugin renewals and the small changes a live shop generates every month.

That range is wider than for a brochure site because the failure cost is higher. A corporate site that breaks is embarrassing. A checkout that breaks is revenue leaving in real time, and it usually breaks on the day of your biggest campaign.

The money is worth it for a specific reason. According to Baymard Institute's meta-analysis of 50 studies, the documented average cart abandonment rate is 70.22 %, and a meaningful share of that is friction rather than intent. Fixing friction is ongoing work, not a launch task.

Ask for the quote to be broken into build, content, integrations and running costs before you compare anything. Two agencies quoting €12,000 and €22,000 are usually not quoting the same project, and the difference is almost always in which of those four the cheaper one left out.

Then ask three questions that cost nothing and reveal everything. Who writes the product descriptions? What happens if the ERP integration takes twice as long as planned? What does month thirteen cost, when the first year of included support ends?

An agency that answers all three in writing is quoting a project. One that answers with a single number and a discount if you sign this week is quoting a hope. Our corporate website pricing is published for the same reason: a price that only arrives after a discovery call is a price designed to be anchored.

Money in ecommerce projects gets wasted in three consistent places, and none of them are the design. First, a platform chosen for features the shop will not use for three years, paid for monthly starting now. Second, a custom checkout built to solve a problem the standard checkout already solves, which then has to be maintained forever.

Third, and most expensive, launching with product content nobody wrote properly. Thin descriptions and borrowed manufacturer copy make the shop invisible in search and unconvincing to the visitors who do arrive, and rewriting 400 products after launch costs more than writing them once.

A cheaper build with real product content will outperform an expensive build with placeholder text every time. If the budget is tight, move money from the build into the catalogue, not the other way round. Our guide to what a website really costs works through the same trade-off for non-commerce sites.