Shopify Apps vs. Your Own Automation: The 500-Euro Trap

No store owner deliberately decides to spend 500 euros a month on apps. They decide twelve times to spend 39 euros.

That is exactly how it went for us. At nano, mate and MUSTAX we ended up with an app stack nobody had an overview of. The total only became visible when I pulled the subscription invoices out of Shopify into a spreadsheet. After that we cleaned up, though not everywhere. What we kept and what we built ourselves is what I explain here, with real numbers.

Why the bill hides

Shopify apps are built so the costs grow quietly. Three reasons.

The individual price looks harmless. 19 euros, 29 euros, 49 euros. Every single decision feels right. Nobody sees the total, because it is not printed on any one page.

Prices scale with you. Many apps bill by orders, contacts or emails. What costs 29 euros at 300 orders a month suddenly costs 199 euros at 3,000 orders. Exactly at the moment when you actually need margin.

There are several stores. Three brands times one app stack means three invoices. For us that tripled the total without adding a single new feature.

There is also a point that never shows up in euros: every app sits in the storefront or the checkout and costs load time. Ten apps are ten scripts. That is a price your conversion rate pays, and your credit card never sees it.

The 500-euro trap, worked through

Here is what a typical stack looks like at a growing store. The prices are market ballparks, not individual quotes.

Even at the low end you land at around 360 euros a month. In the middle it is clearly over 500 euros. Per store.

For comparison, our side of the bill: our self-hosted n8n server costs around 15 euros a month, no matter how many workflows run on it. AI usage for customer service is around 50 euros a month, and that covers all three brands together. The full breakdown is in What automation really costs.

When an app is exactly right

I am not against apps. We pay for a few of them on purpose. Four cases where buying clearly wins.

The problem is standard and solved. Reviews, cookie banners, language switchers. Nobody is going to build something better there, and the price is low.

It is about regulation. Tax, invoicing, data protection. Here you want somebody else to keep up with the law changes. That is money very well spent.

You need it this week. An app is set up in an hour. Your own workflow takes days. When speed counts, the app is the right answer, even if it gets more expensive long term.

You are still testing. As long as you do not know whether a process will stick around, build nothing. At MUSTAX we worked with an app for months just to understand what we actually needed. Without that time we would have built the wrong thing.

The rule of thumb behind it: buy while you are learning. Build once you know.

When building your own gets cheaper

The tipping point is rarely price alone. There are four signals, and they usually show up together.

Signal 1: you pay for 20 % of the feature set. We had a fulfillment app at around 200 euros a month that did about 70 % of what we needed. We covered the rest by hand. So we paid for a tool and did the work anyway.

Signal 2: the app dictates your process. Once you adapt your operations to a piece of software instead of the other way around, you lose exactly what sets you apart from competitors.

Signal 3: you pay several times for the same thing. Three brands, three subscriptions, identical logic. One workflow you build yourself serves all three at no extra cost. That was the strongest argument on our side.

Signal 4: the bill grows faster than revenue. If your app cost per order is rising, you are working against your own scaling.

At mate we built order processing ourselves for those reasons. Two days of development, no subscription costs afterwards. Order processing dropped from 4 hours a day to 15 minutes. The app was never the point, the precise connection to our fulfiller was. The long version of that decision is in Why we build our own tools.

The middle road many people miss

Between app and custom build sits Shopify Flow. It is included in your plan and does more than most people use: tagging orders, warning on risky cases, triggering notifications.

For simple if-then rules you need neither an app nor a server. We wrote down where Flow stops and external automation starts in Shopify Flows vs. custom automation. Check that first, before you buy or build anything. It is the cheapest step of all.

The maintenance cost

This is where build-vs-buy articles fudge the most, so here are the uncomfortable numbers.

A workflow we build ourselves costs us roughly 1 to 2 hours of upkeep a month. For a serious setup across several systems it is more like 2 to 4 hours. Why: APIs change, fulfillers switch formats, Shopify moves versions forward.

Do that math properly. At a 100-euro founder hourly rate, 3 hours of maintenance is around 300 euros a month in time value. Suddenly an app at 150 euros looks cheap.

Two more items that almost nobody plans for:

• Monitoring. Your own workflow dying at night will not call you. On our side every error goes out as a Slack message. Without that you are flying blind.
• Key person risk. If only one person understands the setup, you have a problem the moment that person is out. That is why we document every setup on a single page.

Building your own has a price too. It is paid differently, in time rather than in subscriptions.

How to review your stack in an hour

Something you can do today.

1. List every app subscription with its price. From the Shopify billing page, not from memory. The total at the bottom is usually the surprise.

2. Mark every app you have not opened in the last 30 days. For us that was four. We cancelled three of them and nobody noticed.

3. Sort by price rather than by importance. Take the most expensive app and write down in one sentence what it does.

4. Ask about the most expensive one: what percentage do we actually use? Under half is a candidate for building your own.

5. Check whether Shopify Flow covers the case. That costs you nothing but 20 minutes.

For us this exercise freed up well over 200 euros a month on the first pass, without writing a line of code. Cancelling is the cheapest optimisation step there is.

The limits of this math

So this does not read like an ad for building your own, three cases where it does not add up.

You have nobody to maintain it. Without capacity for the monthly upkeep, every tool you build turns into a ruin. Then the app is the right choice, even at 300 euros.

Your volume is small. With a handful of orders a day, almost no automation pays off. What you save in subscriptions you pay three times over in build time.

The process is still changing. Automate a moving target and you build it twice. Stabilise first, then build.

And one point in our own interest: an agency is also a cost item with a maintenance component. If somebody sells you building your own as free, their math is wrong.

Conclusion

The 500-euro trap is a trap of attention. The price of any single item never hurts enough for anyone to review the app stack.

The decision is simpler than it sounds. Buy standard, buy regulation, buy speed. Build everything that sets you apart from competitors, everything you need several times over and everything where you only use a fraction. And check whether Shopify Flow is enough before you do either.

For us this exact sorting is the reason we run 3 brands with 2 people and still have 33 to 46 hours a week for other things.

If you want to run the numbers on your app stack without anyone selling you something, drop Flowhouse a line. Sometimes our answer is simply that you should cancel two apps and leave it at that.