Shopify or a Custom Build? How to Decide

Every online store project starts with this question and it usually gets answered by whoever is in the room rather than by the business. Someone who has used a hosted platform recommends one. A developer recommends the other. Both are describing their own experience, which is not the same as your situation.

Here is what the decision actually turns on.

What each one really is

A hosted platform is a service you rent. Someone else handles hosting, security updates, payment infrastructure and uptime. You configure rather than build. You are limited to what the platform allows and you pay ongoing fees, often including a slice of revenue.

A custom build is software written for you. No monthly platform fee, no transaction cut, no limits beyond what can be built. In exchange it costs more to start, takes longer, and maintenance is now your responsibility.

Neither is a better technology. They are different trade-offs.

Where a hosted platform clearly wins

You want to be selling within weeks, not months.

Your catalogue is standard. Products with variants, prices, stock counts, a normal checkout.

You do not have a developer and do not want one. Your own team needs to add products, change prices and run promotions without booking help.

You are testing whether the business works at all. Spending heavily on custom software before you know people will buy is an expensive way to learn.

Most stores fit this description, and for most stores the honest recommendation is the hosted platform.

Where a custom build starts to make sense

Your product does not fit the standard shape. Made to order, priced by dimension, configured before purchase, sold as a service with a schedule attached.

You have a system the store must talk to. Stock, accounting, a warehouse process, a supplier feed. Platforms integrate, but only in the ways they have decided to support.

Volume is high enough that percentage fees hurt. On a hosted platform your costs scale with revenue. At some point the annual figure buys a lot of development.

Your process is your advantage. If the thing that makes you better than competitors is how ordering works, and the platform will not let you build it, the platform is now capping you.

The app trap

Hosted platforms have app stores, and every gap has an app that fills it. Individually they look cheap. Together they are a second subscription bill and a performance problem.

Fifteen apps is not unusual on a store that has been running a couple of years. Each one adds code to every page and a monthly charge. It is also a dependency. When one is abandoned by its developer, you find out during an outage.

Before you commit, list the apps you would need on day one. If the list is long, the platform is a worse fit than it looked.

Is there an SEO difference?

Set up properly, both work fine. The difference shows in how much control you have.

Hosted platforms usually fix the URL structure. Product addresses follow a set pattern and you cannot change it. For most businesses this is irrelevant. If you want a specific structure, it is a constraint.

Page speed on hosted platforms is decent by default and degrades as themes and apps pile up.

On a custom build you control both, which is an advantage and a responsibility. A badly built custom store performs worse than a well configured hosted one.

Fees you should actually add up

The monthly platform fee is the number people compare, and it is the smallest part.

Add the transaction percentage on top of your payment processor's cut. Add the apps. Add the theme if it is a paid one. Add anything you pay someone to configure.

Do the same for a custom build. No platform fee and no revenue cut, but hosting, maintenance and the developer time for changes are real and recurring.

Run both totals at your current revenue and at three times your current revenue. The crossover point is often closer than people expect, and it is the only honest way to compare the two.

What happens when the platform changes its mind

You are renting, and the landlord sets the rules. Pricing changes, features get deprecated, an app you rely on gets bought and shut down.

None of this is malicious and most of it is manageable. But it is worth knowing that part of your operation is decided by someone else's roadmap.

On a custom build nothing changes unless you change it. That cuts both ways. No surprises, and no free improvements either.

Who runs it day to day

On a hosted platform your marketing team can add products, launch campaigns and edit pages without a developer. That is a real operational saving and it repeats every week.

On a custom build this depends entirely on how good the admin panel is. If it is weak, every small change goes through a developer and you have bought a bottleneck.

If you go custom, discuss the admin panel during the quote, not after. Establish who can do what without help.

Moving from hosted to custom later

It is possible but it is a project, not a switch. Three things move.

Product and customer data exports cleanly. This is usually the easy part.

Order history is harder. If you want past orders visible in the new system, the mapping rules have to be decided. Sometimes leaving the old system as a read-only archive is the practical answer.

Addresses are the critical part. If product and category URLs change, every old one needs a permanent redirect to its replacement. Skip this and everything you built up in search resets to zero.

The question that usually settles it

Not "which is better" but "what is the cost of being wrong".

Starting on a hosted platform and outgrowing it means a migration in two or three years, by which point you know exactly what you need and can afford it.

Starting custom and discovering the business does not work means you spent months and a lot of money learning something a cheaper route would have taught you in weeks.

The asymmetry points one way for new businesses and the other way for established ones with a process that already works.

What we have seen

We have built both and the pattern is consistent. Businesses that start hosted and migrate later rarely regret the order. Businesses that start custom before they have proven demand often build features nobody uses.

The exception is when the product genuinely does not fit a standard catalogue. In that case the platform fights you from week one and no amount of configuration resolves it.

How to decide this week

Write down what has to be true on launch day. Not the wish list, the actual requirements.

Check each item against what a hosted platform does out of the box. If everything is covered, use it and spend the difference on getting customers.

If two or three core requirements need workarounds, price a custom build and compare properly. A workaround you have to maintain forever is more expensive than it looks on day one.