Own platform or rented? When moving your e-shop actually pays off

Most e-shops belong on a rented platform. Shopify, Shoptet and their peers solve for a few tens of euros a month what would cost tens of thousands to build — and they do it well. When someone offers you a custom platform at the first meeting without asking about your revenue and your processes, they are selling their capacity, not your solution.

But there are situations where the monthly fee stops being a service and becomes a ceiling. This article is about how to recognise them — and what it actually costs.

When is a rented platform enough?

Almost always at the start, and for a long time after. Specifically when most of this holds:

• you sell standard products in a standard way,
• an order follows the usual path: cart, shipping, payment, invoice,
• ready-made connections to stock, accounting and couriers are enough,
• the changes you want can be made in the admin or through an app from their store.

While that holds, migrating gives you nothing you do not already have. It only gives you cost and risk.

How do you know you have hit the ceiling?

1. You pay for features three times. The base plan, plus three to five paid add-ons, plus a developer gluing them together because they do not talk to each other. Once monthly platform and add-on costs pass a thousand euros, it stops being the cheap option and becomes a lease.

2. Your process does not fit the platform. B2B pricing per customer, approving an order before dispatch, a product configurator, a booking with a time slot, a mix of goods and services on one order. These are things rented platforms either cannot do, or do through a workaround that breaks with every update.

3. The integration matters more than the shop. You have an ERP, production or warehouse that is the source of truth, and the e-shop is its shop window. A rented platform gives you an API, but not control over when things recalculate and what happens when a sync fails at three in the morning.

4. You cannot get at your own data. Customers, orders and history sit in a database you can only reach through an export somebody enabled for you. That is not a technical problem until you want to leave — at which point it is the only problem.

One signal is not enough. Two or more at once is a serious indication.

What does an own platform actually involve?

It is not “the same thing, but ours”. It is a system that has to do everything a rented platform quietly does in the background. When we build an e-shop on our own platform, the scope always includes:

• a catalogue with variants, categories, stock and pricing,
• the order flow from cart to dispatch, including statuses and notifications,
• payments and shipping, including the methods that are common locally,
• invoices and credit notes as PDFs, with a payment QR code,
• discounts, coupons and a loyalty programme,
• reviews, returns and complaints,
• an editorial side: pages, sections, menus, a blog,
• an admin a non-technical person can actually operate,
• a background job queue, so generating invoices and sending bulk email does not hold up a customer at the checkout.

That list cannot be shortened with “we do not need it yet”. You need most of it the moment you start selling.

That is why we maintain our own platform and build on it repeatedly. It is not a template — every e-shop has its own processes — but a shared foundation that has already been through production. The price difference between “building from zero” and “building on a proven base” is decisive, and it is the only way a custom platform can be done for sensible money.

What gets underestimated during migration?

Redirects. Old product and category URLs must lead to the new ones. Without a map of 301 redirects you lose search positions built over years, and you will not win them back in a week.

Order history. Customers expect to log in and see what they bought and when. Migrating history is dull work that gets left out of budgets, and if it is skipped it shows up immediately in support.

Content that is not a product. Descriptions, the blog, terms and conditions, FAQs, images. There is usually more of it than anyone remembers.

People. Your back office knows the old admin by heart. The new one is different, even when it is better. Training and running both systems in parallel for a transition period belong in the plan.

How do you decide?

Do not start with “own or rented”. Start with what is stopping you from selling more today. If the answer is “the platform cannot do it and never will”, the decision is clear. If the answer is “we do not have enough traffic” or “our product photos are bad”, migration will not help and the money belongs elsewhere.

If you are not sure, the mini audit is free and it ends with an answer even when that answer is “stay where you are”.