Ecommerce Development Case Study: Meeshop, and When Custom Ecommerce Pays Off

This ecommerce development case covers my work for Meeshop, a Danish platform that lets businesses build and run their own online stores. In short: [PLACEHOLDER: one sentence on what you built or extended for Meeshop, and what it changed for the platform and its merchants]. It also answers a bigger question: when is custom ecommerce worth paying for, and when should you just pick an off-the-shelf platform?

[PLACEHOLDER: Confirm written permission from Meeshop for name, logo, numbers and quote, and confirm the description of Meeshop. Without permission, anonymize the case and change title and slug.]

The short answer: the project at a glance

[PLACEHOLDER: 2-3 sentences in your own words on the starting point, what was built and what it meant for Meeshop.]

For more on product decisions and priorities, read from idea to SaaS: choices and priorities.

The brief: what Meeshop needed

Meeshop sells a hosted webshop solution where the merchant designs the store in a design editor and plugs in the services the business needs. Its site lists MobilePay, QuickPay and Stripe for payments, Shipmondo for shipping and Dinero for accounting: mostly Danish providers alongside one international one. It also says the platform can handle over 1,000,000 products per store, and that merchants who need a specific integration can have one built.

My advice on any ecommerce project: start with the money and the orders, not the design. What happens between the customer clicking "Pay" and the parcel being shipped and booked in the accounts, and where can that chain break?

[PLACEHOLDER: The starting point. What was the need or problem, and how was it handled before you got involved?]

[PLACEHOLDER: How you got involved, what you were responsible for, and what the first delivery had to do. First-person prose.]

What a multi-store ecommerce platform has to handle

A store for one business is a contained system. A platform where many stores share one codebase raises the bar:

• Stores must stay separate. Each store's products, customers, orders and domain stay isolated inside one shared system. This is called multi-tenancy, and a mistake here can show one merchant's data to another.
• Each merchant picks their own providers, each with their own API keys and contracts. Sell across Europe and the list grows, because shoppers in different countries often prefer different payment methods.
• Every release hits everyone. A new version of the platform goes out to all stores at once, including the one running a big campaign that day.

Two EU rules also end up as platform features rather than items on each merchant's to-do list. Once cross-border sales to consumers in other EU countries pass €10,000 a year, VAT on goods is generally due in the buyer's country, which sellers can declare in one place through the EU One Stop Shop (since 1 July 2021). So the platform needs VAT rates per destination country. And the European Accessibility Act covers ecommerce services provided to consumers after 28 June 2025, with an exemption in Article 4 for microenterprises providing services. Templates, the editor and checkout decide most of that, and merchants can't fix them alone. That's my reading as a developer, not legal or tax advice.

[PLACEHOLDER: Which of these demands mattered most in your work for Meeshop?]

Four decisions that shape ecommerce software

These four come up in almost every ecommerce build. For each, here's why it matters and what Meeshop chose.

An order must never disappear

Payment happens at the provider, and the store finds out afterwards through a webhook (an automatic message from the provider's system to yours). Stripe's documentation is explicit that an endpoint may receive the same event more than once, that delivery order isn't guaranteed, and that failed deliveries in live mode are retried for up to three days.

So every order needs clear states (created, paid, shipped, refunded), and every message has to be safe to process twice. Otherwise the customer gets two confirmation emails, stock drops twice, or two shipping labels get booked. The same principle runs through my case study on telecom and fiber web projects, where an order also can't go missing.

[PLACEHOLDER: How payments and the order flow were handled at Meeshop, and whether a particular provider caused specific problems.]

Integrations behind one interface

When a platform talks to several payment providers, carriers and accounting tools, the rest of the code shouldn't care which one is in use. I recommend one shared interface where each integration is an adapter: the platform asks to "create payment" or "book shipment", and the adapter handles the specific provider. Adding a provider for a new market then doesn't mean rewriting the order flow.

A failure in one integration also shouldn't take down the rest. If the accounting system is down, the order still goes through, and the bookkeeping waits in a queue.

[PLACEHOLDER: Which integrations you worked on, and how they were built.]

Releasing without disrupting merchants

On a platform, you can't ship a new version and see what happens. That calls for a staging environment (test setup) that mirrors production, automated tests for checkout and the order flow, and new features switched on for a few stores before everyone. Big changes don't belong in the week before Black Friday.

For the same challenge in a different product with daily users, see the case on building features for a growing employee platform.

[PLACEHOLDER: How testing and releases worked at Meeshop.]

Speed as catalogs grow

Slow product pages only show up once a store has a lot of products. Imports, image processing and stock updates belong in a background queue so shoppers never wait for them, and search across large catalogs usually needs a search index rather than plain database queries.

[PLACEHOLDER: Whether speed or large catalogs were part of the brief, and what was done.]

Results and what I'd do differently

I judge ecommerce software on two things: orders go through every single time, and merchants can manage without contacting support.

[PLACEHOLDER: What was delivered, and when did it go live?]

[PLACEHOLDER: Measurable results Meeshop has approved, e.g. stores, orders, speed or fewer support tickets.]

[PLACEHOLDER: Optional quote from Meeshop with name and title, only with written permission.]

[PLACEHOLDER: One or two things you would do differently today, and why. Prose, including what went wrong.]

When custom ecommerce pays off

Meeshop is itself a good argument against building your own store from scratch. Off-the-shelf platforms have already solved the cart, payments, VAT, shipping, discount codes and order emails, and someone else keeps them maintained. For most merchants, the right move is to pick one and spend the money on stock and marketing.

My rule of thumb has three tiers:

1. If you sell standard products to consumers, use an off-the-shelf platform.
2. If the store has to talk to your ERP, accounting or warehouse system, or needs a special ordering flow, build an integration or an app on top of the platform.
3. If commerce is the product itself (a webshop platform, a B2B ordering portal with customer-specific agreements, or sales tied closely to bookings or subscriptions), custom development can pay off.

Selling in several EU markets doesn't automatically push you to tier three. A standard platform plus a good integration often handles multiple currencies and VAT rules well. For fees and costs side by side, see Shopify vs WooCommerce vs custom.

When not to hire a developer like me

If you just need a store live with a decent theme and standard payments, a developer who builds custom systems is the wrong hire. It'll be slower and pricier than doing it yourself or hiring a partner who specializes in that platform. And if you don't yet know whether your products sell, test the market on an off-the-shelf store and build your own once you know your numbers and bottlenecks.

Next steps

If you're weighing an ecommerce setup that needs more than a standard platform, start by describing how you sell and which systems an order passes through. You don't need a full specification. Larger projects with me start with a paid, fixed-price discovery phase where the scope is defined and priced, and you own the code from day one.

You can read how a project with me runs from first call to launch, or see what I offer for custom web apps and platforms.

Frequently asked questions

How much does custom ecommerce development cost?+

It depends on scope, and a serious estimate needs a defined scope first. The biggest cost drivers are the number of integrations, B2B pricing rules and how much data has to move from the old system. An integration on top of an existing platform usually costs a fraction of a full custom build. That's why I start larger projects with a paid, fixed-price discovery phase, so you get a fixed price you can actually decide on.

Can I start on Shopify or WooCommerce and move to custom later?+

Yes, and it's often the smartest order. Make sure from day one that you can export products, customers and order history, and that the domain and payment account are in your company's name. When you switch, redirect old product URLs to the new ones, or you risk losing search rankings. Plan the move outside your busiest season.

How long does it take to build a custom ecommerce platform?+

Think months, not weeks. A first version with catalog, cart, payments, order handling and one integration takes time to build and even more time to test, because mistakes in payments and orders are expensive. Integrations with existing systems and data migration are the parts that most often run long. A discovery phase gives you a realistic timeline before you commit to the full build.

Who owns the code and customer data in a custom store?+

You do. My clients own the code from day one, and it should live in a repository your company controls. Customer and order data are yours too, and as the merchant you're generally the data controller under the GDPR. Accounts with your payment provider, carriers and hosting should also be in your company's name, so you can change developers without starting over.