Comparison
Low-code or custom development: speed now, a ceiling later
Low-code builds in days what classic development takes months to ship — the question is what happens in year two.
Tools of the Airtable, Zapier and Power Apps variety rewrote the rules for internal software: a non-technical person can now wire a form to a database and a set of automated notifications in a few days, on subscriptions of €0–200/month. For validating an idea or digitising a simple process, the speed-to-cost ratio is unbeatable. The ceiling, however, gives no warning — it arrives abruptly: per-user licences explode when the team grows from 5 to 50 people, performance sags under serious data volumes, integrations with market-specific systems — national e-invoicing platforms such as Romania's e-Factura, local inventory packages — run into the limits of the connectors, and the entire logic of your business lives on a platform free to change its pricing or terms at any moment. Custom development starts indicatively at €5,000 and has no ceiling: the asset is yours, and scaling is purely a budget question. The short rule: internal process, under ~20 users, simple flow — low-code, honestly; a product for customers or a company-critical process — custom. And the hybrid route combines the two intelligently.
Let’s talk about your project
Message us on WhatsApp or send an email — you talk directly to a developer.
office@northdan.com · +40 752 070 247
Key takeaways
Where low-code wins
Time from idea to working tool: days, not months. The cost of failure is minimal — if the process turns out to be badly designed, you have lost a subscription worth tens of euros, not a development project worth thousands.
Where custom development wins
No ceiling: performance at any volume, any integration possible, an interface exactly as you want it. The code is your company's asset rather than rent on someone else's platform, and the cost does not climb with each user you add.
The long-term difference
At 50 users, a €20/user/month subscription means €12,000 a year — indefinitely. A €15,000 custom application amortises against that subscription in just over a year, and afterwards costs only its 15–20% annual maintenance.
Where the balance tips
User count is the financial criterion: under 20, low-code licences are negligible; above that, multiply the licences over three years and set the figure against a custom project starting from €5,000. Flow complexity is the technical criterion: simple conditional rules and forms run beautifully on low-code, but company-specific calculation logic, fine-grained access rights or high-volume processing push you into territory where “configuring” is harder than programming would have been.
Integration is the criterion underestimated most often: standard connectors cover the popular global tools, but a local ERP, a national e-invoicing hookup or a niche production system almost always demands real code around the edges. The moment you notice you are paying consultants to work around the platform's limits, the signal is unmistakable — the ceiling is close, and every additional month deepens a dependence you will eventually have to unwind anyway.
The hybrid route: validate cheaply, rewrite what proves its value
The strategy we recommend even though it takes work away from us: build the first version of any internal tool on low-code, at the smallest budget possible. Half of newly digitised processes turn out to be badly thought through in their first version — and it is infinitely cheaper to discover that on a €20/month Airtable than on a €15,000 application.
After 6–12 months of real use, the picture is clear: the tools the team abandoned, you shut down without regret; the ones that became critical — the ones operations cannot run without — you rewrite in code, with specifications extracted from real behaviour instead of assumptions. That way you pay for development only on software with demonstrated value, and the project's risk drops dramatically.
Frequently asked questions
What happens to the data if I abandon the low-code platform?
The raw data exports almost always, via CSV or an API — but the logic does not: automations, formulas and approval flows export nowhere, and you rebuild them from zero on the new system. Which is why you should document the flows as you build them, not at the end.
Can I build a customer-facing product on low-code?
Technically yes, and for a pilot with your first 10–20 customers it is genuinely a good idea. Over the long run it turns fragile: your margin depends on the platform's pricing, personalising the experience is limited, and some platforms' terms restrict commercial use — read them before, not after.
Who in the company can maintain low-code tools?
An organised person from operations, with no technical training — that is precisely the model's strength. The risk is concentration: when every automation lives in one colleague's head, their departure freezes the tools. Insist on at least two people with access and a written description of the flows.
At what budget does classic development of an internal tool start?
Indicatively from €5,000 for a simple internal application — one flow, a few roles, basic reports — and €10,000–30,000 once integrations with inventory or an ERP appear, at real Romanian market rates. Below those sums, low-code remains the right answer almost every time.
Similar pages
Related resources
Let’s talk about your project
Message us on WhatsApp or send an email — you talk directly to a developer.
office@northdan.com · +40 752 070 247