Shopify vs. custom e-commerce: five signals that the platform is holding you back

Shopify is a good product. For 80% of online shops it makes sense for as long as they are in the “launch and grow” phase. The problem starts when a merchant moves into a more specific operation and realises they are paying for a bundle of features they don't use, while paying extra for the ones that matter.

MessyPlay sells art supplies and runs creative courses. They landed on Shopify at launch — made sense at the time. Two years later they were paying around 6 000 CZK per month — Shopify plus nine plug-ins for things the platform couldn't do out-of-the-box (course booking, Czech VAT nuances, supplier integration). We built them a custom system on Azure at 1 200 CZK per month of running cost, with full functionality. 80% savings aren't a marketing trick — they're the normal outcome when monthly fixed costs outgrow the fixed one-off budget of a custom build.

Here are five signals we see with clients over and over that tell you the platform is starting to stand in the way.

1. Plugin bloat — every month you add another dependency

The classic pattern: Shopify itself doesn't cover some specific need, so you buy a plug-in for $29/mo. Then another for $19/mo for something else. A year later you have nine plug-ins, pay an extra $200/mo, and every plug-in is another vendor that can change pricing, abandon the product, or break on the next Shopify upgrade.

The signal isn't the raw count. The signal is when you look at the list and say: “these four plug-ins would need to talk to each other, but they can't.” That's the moment custom code starts to make financial sense — because those four plug-ins each charge separately AND you maintain the workaround that makes them cooperate.

2. API limits bite on the thing that matters most

Shopify has a reasonable API, but specific moves (bulk operations, custom checkout logic, pricing rules for loyalty programs) run into rate limits or missing endpoints. When your competitive edge lives in something the platform can't do and won't learn, you pay the tax.

For MessyPlay it was the supplier integration. Shopify couldn't verify stock live and would accept an order with a delivery promise it couldn't keep. A plug-in for that existed, but it didn't handle a second supplier. Another plug-in handled both suppliers but not course bookings. And so on. In the custom build it was 120 lines of code.

3. Brand and UX compromises that aren't really compromises

“Shopify themes can do anything.” Sure, 70% of the way. But the 10-15% of details that actually move conversion rates and brand perception require either a CSS hack or a paid custom tweak from a developer, who then checks after every Shopify upgrade whether the hack still works.

When your audience expects a specific UX (colour swatches right on the card, fast reorder from “last purchase”, a custom gifting wizard) and the platform only handles half of it, you start paying twice — once to the platform, once to the customizer.

4. Transaction fees and vendor lock-in add up

Shopify charges a transaction fee unless you use Shopify Payments. Shopify Payments is not available in the Czech market, so you either pay the fee (0.5-2%) or move to Shopify Plus which starts at $2 000/mo. For small-to-medium shops that's 0.5-2% of revenue on top of the fixed fee.

It compounds faster than you think. A shop with 5M CZK annual revenue at 1% fee pays 50 000 CZK/year just because Shopify can't support the Czech payment ecosystem. That can fund a proper integration with Stripe/Comgate/GoPay at fixed cost.

5. TCO flips earlier than you expect

Shopify looks cheap at launch because developer time is expensive. But the price of Shopify + plug-ins + customisations scales linearly with sales volume, while a custom build is mostly one-off development cost plus cheap hosting.

Break-even usually lands between month 18 and 36. The more specific your shop is, the sooner. For MessyPlay it hit at month 18 — one-off development 200k CZK, savings 4 800 CZK/month, full payback within ~42 months including the time saved on plug-in maintenance.

A decision framework, not an ideology

These aren't arguments against Shopify. It's a product that makes enormous sense for the first ~2 years, for testing product-market fit, and for shops whose needs fit the mainstream use case. Ignoring it is a mistake on the other side.

But there's an inflection point where the platform that enabled launch starts to slow growth. If you're wondering whether you are there, do the math: (monthly Shopify) + (plug-ins) + (custom tweaks) + (transaction fees) × 24 = your 2-year TCO. Compare that to a rough custom build estimate (architecture + dev-days × rate + hosting × 24). If custom is 30% more expensive — stay. If it's cheaper or comparable — it's time.

For MessyPlay the calculator gave a clear answer. For shops halfway or three-quarters of the way through a similar calculation it comes out differently, and that's fine. These signals aren't here to chase you off the platform — they're here to tell you when it's worth asking the question.