When a custom online store is worth it instead of Shopify

Let me start with the part that doesn't suit me: if you sell products at a single price, with simple variants and normal logistics, Shopify is fine. It's cheap, it works, it's fast, and no custom build will give you an advantage worth the difference. A developer who tells you otherwise without having looked at your catalogue is selling you days.

The point is that there's a zone where the standard platform stops being enough, and recognising it before you start saves a lot.

The four cases where it isn't enough

Prices aren't a number on the product. Different prices per customer, price lists with validity dates, tiered quantity discounts, terms agreed with a specific reseller. It's the typical situation of anyone who also sells B2B, and it doesn't fit the standard model: it fits by force, with apps piled on top of each other and rules written in three different places.

The product has to be configured. Made-to-measure sizes, materials that change the price, invalid combinations to block, a quote that comes out of the configuration. A serious configurator isn't a variant with more options: it's a calculation, and it has to be written.

Availability is a serious matter. If selling something you don't have means an embarrassing phone call and a refund, the quantity shown has to be the real one net of commitments, updated by whoever runs the warehouse. With one warehouse and few lines it's manageable; with multiple locations, batches or expiry dates, it isn't.

The integration is the job, not an accessory. If the value lies in making the store talk to the business system — orders becoming documents, records kept in line, stock in real time — the site is the tip of work that lives elsewhere. In that case the choice of platform matters less than how the link is built.

The hidden cost of apps

The way stores end up unmanageable isn't a decision, it's a series of patches. A function is missing, you install an app. Another one's missing, another app. Each costs little per month and seems reasonable.

At eight apps you have three problems: a subscription that isn't small any more, functions that step on each other, and a system where you control none of the pieces. When something breaks, every vendor's answer is that the problem belongs to another one. It's exactly the scenario you were trying to avoid by choosing a platform «with no development».

How to decide in half an hour

Write down the five rules your store cannot break. Not the wishes: the rules that, if violated, lose you money or customers. «Customer X pays the agreed price.» «Nothing is sold below minimum stock.» «The order has to reach accounting the same day.»

Then check whether they all fit inside the standard model, with no apps. If they do, open the Shopify account and stop thinking about it. If one or two require gymnastics, it's still doable, knowing that those gymnastics will need maintaining. If three out of five don't fit, building costs less than fighting — and you do that calculation once, at the start, instead of every time the store grows.