Shopify or custom: how to actually decide your e-commerce platform

Platform selection is usually made for the wrong reasons — what the last developer knew, what a competitor uses, or what came up first in a search. It deserves better, because it is the decision in an e-commerce project that is most expensive to undo.

Here is the framework we actually use.

Start with your operating model, not your catalogue size

"How many products do you have?" is the wrong opening question. Platforms handle large catalogues fine. What breaks them is logic — pricing that depends on the customer, inventory that lives in another system, fulfilment that splits across suppliers, or a transaction with more than two parties.

So ask instead: how unusual is the way we sell?

Shopify is the right answer more often than agencies admit

If your model is conventional — you own the inventory, price is the same for everyone, one party fulfils, payment is a normal card transaction — Shopify will get you live faster and cost less to run for years. You inherit hosting, security, PCI compliance, checkout optimisation and payment infrastructure maintained by a company with more engineers on checkout alone than most agencies employ in total.

Agencies underrecommend it because a custom build is a larger invoice. That is worth naming plainly.

Choose Shopify when: you are direct-to-consumer, your pricing is uniform, you want to spend your energy on merchandising and marketing, and no critical integration is missing from the app ecosystem.

CMS-based commerce, when content carries the sale

Some businesses sell through explanation. Education, considered health products, technical equipment, brand-led categories where the story does the persuading — here the content and the catalogue have equal weight, and a commerce layer on a strong CMS beats a store with a blog attached.

Choose CMS-based when: the majority of your traffic arrives on content rather than product pages, and the path to purchase runs through explanation.

Custom, when the platform would be fighting you

Custom is right when your operating model cannot be expressed in an off-the-shelf platform without constant workarounds. In practice, that means one of:

• Multi-vendor or marketplace logic — seller onboarding, commission handling, split payouts.
• B2B pricing complexity — negotiated rates, customer-specific catalogues, quote-to-order, credit terms.
• Deep operational integration — where the store must be a live participant in an ERP or manufacturing system rather than exporting orders to it.
• Regulatory or contractual constraints on where data lives and who processes it.

Choose custom when: you have already tried to make a platform fit and are maintaining a growing pile of workarounds — that is the clearest signal there is.

Model the three-year cost, not the build

The build quote is the least useful number in the comparison. Model both options across three years including platform fees, transaction percentages, app subscriptions, hosting, security maintenance, developer time for changes, and the cost of the workarounds you already know you will need.

Platform fees and transaction percentages scale with revenue; custom build costs largely do not. There is a crossover point. Find yours before deciding, rather than discovering it in year two.

The question that settles most cases

If you could not change your platform for three years, which decision would you regret less?

Most businesses answer that quickly, and correctly.