Shopify vs Custom E-commerce: An Honest Comparison

I build custom e-commerce for a living, so you'd expect this post to be a takedown of Shopify. It isn't. Shopify is a genuinely good product, and for a lot of stores it's the right answer. I've told people exactly that, for free, when they came to me asking for a custom build.

But "Shopify is good" and "Shopify is right for you" are different claims. The honest answer depends on your volume, your margins, and how custom your selling experience needs to be. Here's how I think through it.

Where Shopify wins, clearly

Speed to market. You can have a working store this weekend. Products, checkout, taxes, shipping rates - solved out of the box. No custom build gets you live faster, including mine.

The ecosystem. Thousands of apps, thousands of themes, and every 3PL, email platform, and marketing tool ships a Shopify integration on day one. When something breaks, the answer is a support article away.

The checkout itself. Shopify's checkout is battle-tested across millions of stores. Fraud handling, saved payment methods, Shop Pay - that's years of conversion optimisation you get for free.

If you're validating a product, doing under a few thousand euros a month in revenue, or your store is a straightforward "catalogue plus checkout," stop reading and use Shopify. Genuinely. A custom build at that stage is over-engineering, and I'd rather tell you now than invoice you for it.

The creep

The problems with Shopify are rarely visible in month one. They show up gradually, in two forms.

First, transaction economics. The Basic plan runs about €30+ a month, which sounds cheap. But unless you use Shopify Payments, you pay an extra fee on every transaction on top of your payment provider's cut - and Shopify Payments isn't available or ideal in every country or for every business model. Card fees themselves land around 1.5-3% depending on plan and region. On €5,000 a month that's noise. On €50,000 a month, those percentage points are a salary.

Second, app subscriptions. The base platform is deliberately minimal, so real stores accumulate apps: reviews (€15/mo), subscriptions (€40/mo), advanced search (€30/mo), bundles, upsells, loyalty, translations, back-in-stock alerts. I've seen public teardowns of mature Shopify stores paying €300-800 a month in apps alone - each one adding its own scripts to your storefront, which is why so many heavily-apped Shopify themes score poorly on Core Web Vitals despite the platform's solid foundation.

Then there's the ceiling. The moment you want something the platform didn't anticipate - a genuinely custom checkout flow, unusual pricing logic, a product configurator, deep integration with your own backend - you're either paying for Shopify Plus (from roughly €2,000/month) or fighting the platform.

When custom pays off

A custom store - my stack is Next.js, React, and TypeScript with Stripe for payments, PostgreSQL and Prisma behind it, deployed on Vercel - starts at €3,000 and takes about 4-8 weeks. That includes custom checkout and user flows, Stripe integration, an admin dashboard, and A/B testing infrastructure, with three months of post-launch support. Details on the pricing page, and you see a full design mockup before paying anything.

The maths flips in custom's favour when some combination of these is true:

Volume. Stripe charges its processing fee and nothing else - no platform percentage stacked on top, no monthly plan scaling with your ambitions. Hosting on Vercel runs €0-20 a month for most stores. At meaningful revenue, the monthly delta versus a fully-apped Shopify setup often covers the build cost within the first year or two.

Differentiation. If your buying experience is your competitive edge - custom configurators, unusual checkout logic, content and commerce deeply interwoven - then a template platform actively works against you. On custom, the weird feature that makes your store yours is just code.

Performance. Speed is a conversion lever, not a vanity metric. Google's research shows bounce probability rising 32% as load time goes from one to three seconds. A lean Next.js storefront routinely loads in under 1.5 seconds; a Shopify theme carrying fifteen apps often can't. I run my own products on this stack - MyHomeStock (web + iOS), QRify, PDFer with its 31 PDF tools - so the performance claims aren't theoretical, they're pages you can load yourself.

Ownership. Your customer data, your logic, your infrastructure. No platform lock-in, no app that changes pricing on you, no risk of policies shifting under your feet.

Migration paths, both directions

This isn't a marriage. Moving Shopify → custom is well-trodden: products, customers, and order history export cleanly via Shopify's APIs, and the usual approach is to build the new store in parallel, import the data, and switch DNS when it's ready - no downtime, no lost SEO if redirects are done properly. Some stores even go hybrid first: a fast custom Next.js storefront on top of Shopify's backend via their Storefront API, which fixes performance without touching operations.

Going the other way - custom → Shopify - is also fine, and worth naming: if a custom build was the wrong call, you're not trapped. The data exports; nothing about custom code holds your catalogue hostage.

My honest default

Under roughly €10k a month in revenue with a standard catalogue: use Shopify, and revisit in a year. Growing volume, fee-and-app fatigue, or a selling experience that needs to be genuinely yours: custom starts making financial sense fast, and from €3,000 the entry point is lower than most people assume.

If you're weighing this for a real store, tell me your numbers - revenue, current monthly platform costs, what's frustrating you. I'll give you a straight answer, including "stay on Shopify" if that's the truth.