A €2,000 shop and a €50,000 shop sell the same products. Here is what the other €48,000 buys.

Put the two side by side on a phone and most people cannot tell which is which. Both have a product grid, a cart, a checkout and a card form. Both take money. That resemblance is why the question gets asked in the first place — if the cheap one works, what is the expensive one for? The honest answer is that they are not better and worse versions of the same thing. They are solutions to different problems, and the difference between them is almost entirely invisible until a specific threshold gets crossed. This guide names those thresholds, so the decision stops being about taste and starts being about which ones apply to you. The first and largest of them is not a feature at all: at one end you are configuring a platform, at the other you are having software built.

The same shop, at both budgets

Two rows carry most of the gap. The first sets the ceiling on every row beneath it: configured shops do what the platform already does, developed shops do what gets written. The last one does the opposite and does not move at all — accessibility and product-safety obligations apply to a €2,000 shop and a €50,000 shop identically, because budget does not lower a legal floor.

Both shops look the same in a screenshot, and that is the trap

Design is the thing everyone compares, and it is the thing that scales least with money. A capable designer produces a good product page in roughly the same time whether the shop behind it holds 40 products or 40,000, and a modern platform theme starts from a layout that has already been tested against more real shoppers than any of us will ever run an experiment on. What the bottom of the range buys is that theme fitted to your brand — your colours, your type, your photography in the slots it already has — not a layout designed around your products. Past a certain point the visual work is not what a bigger budget is buying either, and any quote that presents it that way is describing the wrong project.

What actually costs money is everything that has no screenshot. A stock sync that does not oversell during a flash sale. A checkout that survives an address format nobody on the team has seen before. A returns flow that puts the item back into inventory and the money back on the card without a human copying numbers between two tabs. None of that appears in a mockup, all of it appears in the invoice, and the businesses that get burned are the ones that compared mockups.

This is also why the honest version of this conversation starts with your operations rather than your taste. Ask what a shop has to talk to, how many decisions a day it has to get right on its own, and what a wrong one costs. The number falls out of those answers, and it is remarkably stable once you have them.

Configured or developed — that is the actual line

Everything else in this guide follows from one distinction, and it is not a feature anyone can point at in a demo. A shop at the bottom of the range is configured. A platform is installed, a theme somebody else designed is fitted to your brand, payment and tax settings are filled in, and products are entered. No code is written for you. That is not a criticism — it is the whole reason the work is fast, and the only reason the price can be fixed in advance at all.

A shop at the top is developed. The templates, the content model, the pricing rules, the checkout steps and the connections to whatever else you run are written for your business specifically, on a custom or headless foundation rather than inside somebody else’s admin panel. That is what our e-commerce and B2B platforms work is: catalogue structure and pricing logic built around how you actually sell, instead of bent to fit whatever a plugin was willing to do.

The consequence is a ceiling, and it is the single most useful thing to understand before asking anyone for a quote. A configured shop does what the platform and its plugins already do, and nothing else — when you reach that edge, the honest answer is “no”, not “spend more”. A developed shop does whatever can be built, which is why the top of the range has no natural limit and why a serious quote for one begins with discovery rather than a price list.

Almost every unhappy e-commerce project we get asked to rescue is the same story: a configured shop sold as though it had no ceiling. The build was fine. The expectations were sold against a different product.

Checkout is where the money already is

Roughly seven in ten carts are abandoned. The Baymard Institute keeps a running average across 50 separate studies and puts it at 70.22%, and that figure has barely moved in a decade — it is a property of how people shop, not a symptom of a broken shop. Some of it is unreachable: people browse, compare and save things for later, and no amount of engineering converts a window shopper.

The reachable part is large enough to be the whole argument. Baymard’s own checkout research finds that a large e-commerce site can gain a 35.26% increase in conversion rate through checkout design alone, and puts the recoverable total across US and EU e-commerce at around $260 billion. The most quoted number underneath that is simply field count: the average checkout displays 23.48 form elements by default where 12 to 14 would do, and asks 14.88 questions where 7 or 8 would.

Here is the part that decides your budget. A stock platform checkout is already close to that ideal, because the platform optimised it against traffic volumes no individual merchant can match. At €2,000 you inherit that work for free, and the correct move is to change as little of it as possible. At €50,000 you are paying to go past it — cutting fields that are only there because a system upstream demands them, handling address formats per market, choosing a payment mix per country, and testing all of it against your own funnel rather than someone else’s average. That is worth doing when a percentage point of conversion is worth more than the work costs, and it is worth nothing at all before then.

The trap in between is a custom checkout built for aesthetic reasons on a small shop. It reliably converts worse than the default it replaced, because it discarded a decade of platform testing to gain a rounded corner. If your budget is at the lower end, spend it on photography, copy and delivery promises, and leave the checkout alone.

Threshold one: how deep the catalogue goes

Catalogue size is the first place a cheap shop genuinely stops working, and it does not degrade gracefully — it is fine, then it is not. A few hundred products with straightforward variants sit comfortably inside any platform’s built-in tooling. A few thousand, with variant matrices, bundles, kits or per-customer pricing, do not, and the failure shows up in three places at once: the admin becomes unusable for whoever manages products, built-in search stops surfacing the right things, and page performance falls off as category pages try to render more than they were designed to.

Search is the part most often underestimated. On a large catalogue, on-site search is where the buyers are — people who use it are already past deciding whether to buy and are only deciding what — and the built-in search in most platforms is a simple keyword match with no typo tolerance, no synonyms and no merchandising control. Replacing it with a dedicated search service is one of the highest-return line items in a large build, and one of the most pointless in a small one, where a visitor can simply look at the category page and see everything you sell.

The threshold question is not “how many products do I have” but “can one person keep the catalogue correct”. While the answer is yes, you are in €2,000 territory no matter how ambitious the roadmap sounds. Once it takes a process, a spreadsheet import and a second pair of eyes, the shop has become a system that needs building rather than configuring.

Threshold two: how many systems have to agree with each other

A small shop has one source of truth, and it is the shop. Stock, prices and orders live there, someone edits them there, and if two numbers disagree there is exactly one place to look. That single property is most of why a €2,000 build is cheap, and it is the first thing a growing business loses.

The moment stock lives in a warehouse system, prices come out of an ERP and invoices belong to accounting, the shop stops being the source of truth and becomes one participant in a conversation between systems that were never designed to talk. The cost is not the connection itself — plenty of these have an off-the-shelf connector. The cost is the disagreements: what happens when the ERP is unreachable during a sale, what wins when two systems both changed the same price, how an order that was paid but not written to the warehouse gets found. Every one of those has to be decided, built and tested, and none of it is visible to a shopper.

This is the single biggest driver of the gap between the two budgets, and it is also the one businesses recognise instantly once it is named. If you already run an ERP, a WMS or a warehouse team, you are not buying a shop, you are buying an integration project with a shop attached. Our API integrations work is scoped around that reality, and CRM Integration at €499 is the smallest honest version of it — one system, one direction, properly done.

If, on the other hand, everything already lives in one place and you are happy for it to keep living there, do not build for a future you have not reached. WooCommerce Store is €1,499 and covers a real catalogue with payments and a tested checkout flow; Shopify Store is €999 if you would rather rent the platform and never think about the server. Both of those are the right answer far more often than the industry admits.

Threshold three: how many countries you sell into

Adding a language to a shop is cheap. Adding a market is not, and the two get confused constantly. A second language is translated content against the same catalogue, the same prices and the same tax treatment. A second market is a different price list, a different VAT rate, a different payment mix, different delivery promises, different legal pages and a different returns address — and it multiplies through every product and every category page you have.

The EU rules have a specific number in them that makes this concrete. Below €10,000 a year in combined cross-border B2C sales, you may keep charging your own country’s VAT. Above it, you charge the customer’s country rate, and you either register in each member state or file through the One Stop Shop. That threshold has covered intra-Community distance sales of goods since 1 July 2021, and it counts everything together rather than per country. For goods imported from outside the EU there is a second scheme, IOSS, for consignments not exceeding €150.

The practical consequence is that international is a step change rather than a slope. A shop that crosses €10,000 in cross-border sales needs per-market VAT handling, per-locale legal pages and per-market pricing, and retro-fitting that onto a single-market build is more expensive than having planned for it. Multilingual E-commerce is €2,999 and is scoped exactly around that step — a translated catalogue, localised checkout and confirmation mails, and hreflang across product and category pages. The architecture side of the same problem is covered in multilingual SEO architecture.

Threshold four: what an hour of downtime costs

This is the least discussed line in any e-commerce quote and often the largest. At €2,000, the recovery plan is that you notice something is wrong, you contact support, and the shop comes back within the day. For most small merchants that is genuinely acceptable — a quiet Tuesday afternoon costs very little, and paying to eliminate it would cost more than the outage.

At the other end, an hour of a failed checkout during a campaign is a number somebody can calculate, and once it can be calculated it justifies spending against it. What the money buys is not magic: monitoring that alerts before a customer does, a deployment process that can be reversed in minutes, a staging environment that resembles production closely enough to be worth testing on, and a named person who is accountable during the window. None of that is exciting, all of it is expensive, and it is bought entirely on the arithmetic of what an hour is worth.

Work out that number before asking for a quote. It is the single fastest way to find out which shop you are actually buying, and it settles more arguments than any feature list. If the honest answer is “a few hundred euros”, stop reading price ranges and buy the small one.

Compliance is a floor, not a tier

The row in the table above that does not change with budget is the important one. Two EU rules apply to a small shop and a large one in exactly the same way, and neither of them cares what you paid.

The European Accessibility Act has been enforceable since 28 June 2025 and names e-commerce explicitly. It is enforced against EN 301 549, which is built on WCAG 2.1 AA, and it reaches businesses outside the EU that sell to EU consumers. A themed shop is not automatically conformant, and the parts that most often fail — keyboard access through the checkout, focus visibility, form labels, contrast in the states nobody screenshots — are the parts a cheap build is least likely to have checked. Our accessibility work covers the audit and the remediation, and the WCAG and EAA guide sets out what conformance actually requires.

The General Product Safety Regulation has applied since 13 December 2024 and reaches the product page itself. For anything sold at a distance, the offer must carry the manufacturer or their representative’s details, a clear description of the product, and any warning or safety information — the same information a shopper would find on the box in a physical shop. That is a data-model requirement rather than a design one: those fields have to exist on every product, be populated by whoever adds products, and render on the page. On a shop with 60 products it is an afternoon. On a shop with 6,000 it is a project, and it is exactly the sort of thing that gets discovered after launch.

The platform decides your performance before you write any code

Speed is the one area where the cheap option is frequently the faster one, which surprises people who assume budget and performance move together. The HTTP Archive’s 2025 Web Almanac measured Core Web Vitals pass rates across real e-commerce sites by platform: Shopify passed on 76% of desktop sites, Wix eCommerce on 70% and Squarespace Commerce on 69%, while PrestaShop managed 54%, Magento 36% and WooCommerce 33%. Mobile is consistently worse than desktop across all of them.

Read that carefully, because it is easy to draw the wrong conclusion. It does not mean WooCommerce is slow — it means the average WooCommerce site is, because it is assembled from plugins by whoever could be afforded, while a hosted platform constrains what anyone can do to it. A well-built WooCommerce shop outperforms most Shopify shops. A cheap one almost never does. What you are seeing in those numbers is the difference between a platform that has removed your ability to make things worse and one that has not.

The budget consequence is direct. If you are spending at the low end and cannot afford performance work, choose the platform that protects you from yourself. If you are spending at the high end, the platform stops being the constraint and the discipline becomes yours to keep — which is what Speed Optimisation at €799 and the Core Web Vitals guide are for.

What we would actually do with each budget

With €2,000: pick the platform that matches who will run the shop, use the stock checkout, put your own product photography and delivery copy ahead of layout, connect analytics, and check accessibility before launch rather than after. Then be equally clear about what that budget does not reach — no custom fields, no integration with anything you already run, no change to how checkout behaves, and a catalogue in the dozens rather than the thousands. Inside those limits it sells the same products as the expensive one. Outside them it does not bend, and no amount of goodwill from whoever built it changes that.

With €50,000: expect roughly a third to go on integration work you cannot see, a meaningful share on the catalogue and search layer, and only a modest share on visual design. Insist that the quote separates discovery, integrations, data migration, QA and post-launch support into their own lines — a single number for “e-commerce website” hides which of those is being skipped, and the one being skipped is almost always QA or migration.

The awkward middle is the range to be most careful about. A €12,000 to €20,000 budget is enough to start behaving like a large project and not enough to finish like one, and it is where half-built integrations and abandoned custom checkouts live. If you land there, our advice is consistent: buy the small shop properly, keep the difference, and spend it on the one threshold you have actually crossed rather than on a preview of all four.

We are a small studio with fixed public prices, which means we do not have a five-figure e-commerce quota to fill. Telling someone their €50,000 brief is a €3,000 project costs us the bigger invoice, and it has been the right call every time we have made it.

How to read a quote that lands between the two

Most real quotes arrive somewhere in the middle, and the way to judge one is not by its total. Ask for the same brief priced twice: once on a hosted platform with the stock checkout, and once the way the agency wants to build it. The gap between those two numbers is the actual proposal — it is, almost exactly, the price of development over configuration — and it becomes very easy to ask what each part of it buys.

Then check three things in the breakdown. Whether data migration has a line at all — it is the most commonly omitted and most reliably underestimated item in an e-commerce project. Whether QA is a line or an assumption. And whether there is a named post-launch window, because the first two weeks of real traffic find things no staging environment does. A quote missing all three is not cheaper than one that includes them; it is the same project with the risk moved onto you.

If you want the arithmetic behind our own prices rather than a range, the price guide sets every package against published agency rates, and how much a website costs covers the drivers on the non-commerce side.

Sources

The abandonment and checkout figures, the platform performance data and both EU rules come from these, checked August 2026. The four thresholds, the budget splits and the read on where the money goes are ours, from client projects.

• Baymard Institute — Cart Abandonment Rate Statistics ↗

  The running average across 50 separate studies: 70.22% of online shopping carts are abandoned, latest data retrieved August 2025. Also the source for the 35.26% conversion gain available from checkout design alone on a large site, and the ~$260 billion recoverable across US and EU e-commerce.

• Baymard Institute — Checkout Usability research ↗

  The field-count figures: an average checkout displays 23.48 form elements by default where 12–14 would be enough, and asks 14.88 questions where 7–8 would do. The basis for treating a stock platform checkout as already close to optimal on a small shop.

• HTTP Archive — Web Almanac 2025, E-commerce chapter ↗

  Core Web Vitals pass rates on real e-commerce sites by platform in 2025: Shopify 76%, Wix eCommerce 70%, Squarespace Commerce 69%, PrestaShop 54%, Magento 36%, WooCommerce 33% on desktop, with mobile consistently lower. Also the platform market shares quoted above.

• EUR-Lex — Directive (EU) 2019/882 (European Accessibility Act) ↗

  Applied by member states since 28 June 2025, naming e-commerce among the covered services, and reaching businesses outside the EU that sell to EU consumers. Enforced against EN 301 549, which is built on WCAG 2.1 AA.

• EUR-Lex — Regulation (EU) 2023/988 (General Product Safety Regulation) ↗

  Applicable since 13 December 2024. For distance selling, the pre-contractual offer must carry the manufacturer or representative’s details, a clear description of the product and any warning or safety information — the same information available in a physical shop.

• European Commission — VAT One Stop Shop ↗

  The €10,000 combined annual threshold below which a seller established in one member state may keep charging its own country’s VAT; above it, the customer’s country rate applies, declared through OSS. Covering intra-Community distance sales of goods since 1 July 2021. The import scheme (IOSS) covers consignments not exceeding €150.