Shopify vs custom e-commerce: a platform or your own online store

Three options, not two

The choice is usually framed as "Shopify or custom", but there are really three families, plus a hybrid:

• Hosted platforms such as Shopify, Wix and many local equivalents. You rent the store: hosting, security, checkout and updates are the platform's job. You pick a theme, add apps and pay a monthly fee, often plus transaction fees.
• Open-source engines you host, such as WooCommerce, OpenCart or Magento. The software is free; hosting, updates, security and customisation are yours, usually through a developer or an agency.
• A custom store: your own storefront, catalogue, checkout, customer accounts and admin, built for how you sell and owned by you.
• Headless: a platform's back end for products, orders and checkout, with a custom-built storefront in front of it.

Each one trades money, speed and control differently.

When a platform is the right answer

If a platform covers your catalogue and checkout, start there. It's cheaper. That's the case when:

• your products fit a standard catalogue: a product, variants, a price, stock;
• buyers decide by looking at photos and a description, not by working something out;
• your checkout is a normal cart with delivery and card payment;
• you need to be selling in weeks, with a small budget;
• you're still testing whether the product sells at all.

For many shops, this is all they'll ever need. There is nothing second-rate about it.

When a custom store pays off

Build your own when the way you sell doesn't fit a template, or when the platform starts costing more than ownership. In practice, the signs look like this:

• Buyers need help to choose. They need an expert's guidance, a selection by use or compatibility, or a calculator that turns a hard decision into a number.
• B2B is part of the business. Customer-specific prices, invoices for companies, repeat orders, credit terms, several buyers in one company account.
• Your integrations are unusual. Accounting, warehouse, ERP or partner systems that no maintained app connects reliably.
• The app bill keeps growing. Each missing feature is a paid add-on, and together they cost more than the plan itself.
• Search traffic matters a lot. You want full control over page structure, speed, structured data and content pages for the questions buyers search for.
• You want your data and logic under your control, not split across a platform and a dozen app vendors.

What each option costs over time

For orientation only, quotes from studios in Ukraine and Eastern Europe in 2026 look roughly like this:

If you set up a hosted store yourself, the subscription may be your only cost at the start. These ranges are rough, and your scope decides the real number. See how much a custom online store costs for the drivers in detail.

The costs people miss

On platforms:

• apps that each look cheap and together don't;
• an extra transaction fee on some platforms when you don't use their own payment service, which may not be available in your country anyway;
• deep changes to checkout may be limited or reserved for higher plans, so check before you commit;
• a theme that fights you every time the business changes;
• the work of moving out, if you ever do.

On custom builds:

• the admin has to be built too, since there's no platform admin to lean on;
• hosting, monitoring and security updates are your responsibility, or your vendor's under a support agreement;
• the first launch takes longer than switching on a theme;
• every feature has to be decided, not just switched on.

A test before you decide: walk the buyer's path

Before comparing price lists, write down how your best customer actually buys, step by step, from the first visit to a repeat order. Then mark each step:

• Standard: browse, look at photos, add to cart, pay, track delivery. Any platform does this well.
• Needs advice: "which product for my case?", "how much will I need?", "is this compatible with what I have?".
• Needs your data: personal prices, stock in a specific warehouse, a delivery date that depends on production.
• Needs a person: a consultation, a quote, an approval.

If almost every step is standard, a platform is the honest answer. If the steps that decide the sale need advice or your data, those are the steps a template can't carry, and they're usually where a custom store earns its cost.

If you choose a platform, choose it carefully

A platform is easier to leave if you check a few things on day one:

• you can export products, customers and orders in full, not just a summary;
• you control page addresses, titles and descriptions, so a future move doesn't cost your rankings;
• the apps you depend on are maintained, and you know what each costs per month;
• the platform has an API, in case you later need a custom storefront or integration;
• the plan's fees and commissions are spelled out, and you know what changes when your turnover grows.

The middle path: headless

In a headless setup the platform keeps the products, orders and checkout, and a custom storefront sits in front. You get a fast, fully designed front with the platform's reliable checkout behind it.

It fits stores that need a distinctive storefront and strong SEO but a fairly standard checkout. It doesn't remove platform fees, and it adds a second system to maintain. If what doesn't fit is the checkout or the B2B logic, headless won't solve it.

An example: TAMIS AGRO

TAMIS AGRO supplies Ukrainian farms with biological products: bio-insecticides, inoculants, micronutrients and organic fertilisers. For many farms biologicals are a new idea, and an agronomist used to chemistry wants to know what replaces what, at what rate and at what cost.

So the store had to sell like a catalogue and advise like an agronomist. The catalogue follows the way an agronomist searches, by category, crop and stock, with the rate, packaging and price on every card. Crop pages match each culture with proven products. A calculator compares chemical and biological protection costs year by year, which turns the case for biologicals into numbers. Customers keep addresses and order history in their profile, and every page ends with a person to talk to: a callback or a free consultation.

None of these pieces is exotic, but together they describe a way of selling that a standard theme doesn't. That's the typical moment a custom store makes sense. Our guide to an online store for agricultural products goes deeper into this niche.

Local details that decide the choice

Before choosing a platform, check that it works with the services your business actually uses:

• Payments: the card acquirers available in your country, invoices for companies and, where relevant, cash on delivery. We connect WayForPay or Stripe for card payments in custom stores; read accepting online payments in Ukraine for the wider picture.
• Fiscal receipts: in Ukraine, ask your accountant whether your online payments need receipts through a software cash register (PRRO), and how the platform handles them.
• Delivery: the carriers your customers expect, such as Nova Poshta, Ukrposhta or Meest in Ukraine, with live rates and branch selection.
• Accounting and stock: whether prices and stock can sync both ways with the system your accountant uses, BAS for example.

If an integration you depend on is missing or poorly maintained, that alone can be a reason to leave a platform later.

Moving from a platform to your own store

Starting on a platform and moving later is a sensible path. A careful move covers:

• the catalogue, with every product, variant and image;
• customers and their order history, so returning buyers can sign in;
• URLs: every old address redirects to its new equivalent, so search rankings carry over;
• content pages and their metadata;
• a test period with real orders before the switch.

At EVA-X we do this as part of building a store: your own store is built first, then the catalogue and customers are moved into it.

A decision checklist

Choose a platform if most of these are true:

• standard catalogue and checkout;
• a small team and budget;
• you need to sell within weeks;
• the integrations you need are available and maintained.

Look at custom development if most of these are true:

• buyers need guidance, calculators or configuration to decide;
• B2B prices, accounts or invoices matter;
• your integrations are unusual;
• platform apps and fees already cost more than the plan;
• organic search is a major channel.

For a standard catalogue and checkout, we usually recommend the AI + DEV format, where AI builds the familiar screens and engineers own payments, integrations and SEO. A one-off promo shop can go AI Build. Large catalogues with complex B2B pricing and integrations deserve DEV. More on B2B in the features that matter in B2B e-commerce.

To see what your store would cost, describe it on the estimate: it lists the modules, the timeline and a price in about 15 seconds, and our team reviews it with you before any contract. Or read about our e-commerce development.