Shopify or a custom store — when each is right

"Should we just build our own store" comes up in almost every ecommerce conversation, usually from someone frustrated by an app limitation or a theme that will not quite do what they want. The frustration is real. The conclusion is usually wrong.

What Shopify is actually selling you

Not a website. A checkout that has processed enough transactions to have found and fixed the failure modes you have not thought of yet — abandoned carts, fraud, tax rules that change by state, a payment that fails halfway through. Building that yourself means rebuilding all of it, quietly, one bug report at a time.

That is what the subscription buys, and it is worth more than the theme.

When Shopify is the right answer

You sell products, at one price, to whoever turns up. This is most small and mid-sized stores, and Shopify was built for exactly this shape.

Your catalogue is large but ordinary. Hundreds or thousands of products with normal variants — size, colour, quantity — are what the platform is designed to hold, whatever the number looks like.

You are not sure yet. Set the store up yourself and sell for a month before spending on anything more. What the store actually needs will be clearer after a month of real orders than after any amount of planning.

When a custom build is worth considering

Selling rules a cart cannot express. Different prices for different customers, credit terms, approval before an order ships, quantities that only make sense in bulk. This is the B2B shape, and it is the clearest genuine case for custom.

The buying flow is not really a cart. A configurator, a quote that has to be approved before checkout, a subscription with rules a standard app cannot enforce. If the checkout itself is where your business is different, a platform built for a generic cart will fight you the whole way.

You are the product, not the store. If what you are building is software that happens to include commerce — a marketplace, a platform other businesses sell through — that is a software project, and it should be scoped as one from the start.

The case against building, stated plainly

A custom store is software you now own, with everything that implies: hosting, security patches, PCI compliance for handling card details, and a checkout that has to keep working every time a payment provider changes something on their end. None of that shows up in a demo. All of it shows up in year two.

"We want it to look exactly like us" is not a reason to build — a theme can be customised further than most businesses try. "We want full control" sounds like a reason until you notice that control also means you now own every bug.

Shopify is not the beginner option that serious businesses grow out of. Most businesses that grow out of it were never the shape it was built for in the first place.

The order that avoids an expensive mistake

Start on the platform, always, even if you suspect you will outgrow it. Run real orders for long enough to know exactly where it stops fitting — not where you imagine it will stop fitting.

If a specific, nameable thing does not fit — a pricing rule, an approval step, a buying flow the cart cannot express — build only that piece and connect it to the platform, rather than replacing the whole thing. The same logic applies to software generally: buy what is standard, build only what is genuinely yours.

Most stores that move to custom too early spend the next two years rebuilding what Shopify was already doing for them, badly, at greater cost, with nobody else's engineering team behind it.