No-code or code: where no-code wins and where the ceiling is
No-code can be a practical first version, internal tool or demand test when its permissions, workflow and costs fit. Compare the actual platform and requirements rather than assuming it is always the right starting point. Code from us starts at EUR 600 to 900 for a single-page landing at step 01, runs EUR 3.500 to 7.500 at step 04 for a web app with logins, a dashboard, an API or payments, and starts from EUR 7.500 for a custom SaaS platform once discovery has fixed the scope.
We make our money on the code side of that sentence, so read what follows with that in mind. We start with the case against hiring us.
Where no-code genuinely wins
By no-code we mean app builders such as Bubble, Glide and Softr, databases with a friendly face such as Airtable, and automation tools such as Zapier and Make. Website builders are a related but different trade.
- Before you have a paying customer, the costliest mistake is building the wrong thing well. A Softr front on an Airtable base tests demand for the price of a subscription.
- An internal tool can suit a visual builder, but it still needs usable screens, correct permissions and reliable data. Check external collaborators, sensitive information and the actual plan limits.
- Glue between tools you already pay for is a Zapier job. A form entry that lands in a spreadsheet and posts a message to your team does not need custom code, and paying for it would waste your money.
- The person who knows the process can build it, without writing a brief or waiting for a developer to learn how your stock or your rota works.
- Changing your mind is cheap. You can rebuild a screen tonight and throw it away tomorrow.
If that list describes where you are, stay there. A no-code tool that works is not a problem to be solved.
Where the ceiling is
It is not one wall. It is a handful of separate limits, and they bite hardest once real customers and real money are involved.
Permissions
Who can see which record is the part of any app that goes wrong quietly. Visual and coded access rules both need testing with different users and roles. Check which automated and manual tests the selected platform supports; the editor type alone does not establish security or testability.
The bill follows usage, not value
No-code plans charge by seats, records, tasks, operations or usage units. That is fair while you are small. The actual bill depends on the plan and its thresholds. Review available workflow and query optimisations, capacity and rate limits before deciding that a rebuild is needed.
You can export the data, not the app
Airtable exports your tables. Bubble does not export your application as code. The logic you spent evenings building lives on the platform, and leaving means rebuilding it somewhere else.
The connector list is the edge of the map
Check the platform's API, authentication, webhook and retry features before concluding that custom code is required. A missing ready-made connector is a constraint to investigate, not proof that the integration is impossible without coding.
Public pages and search
Some app builders support public, indexable pages. Check the actual rendered content, status codes, titles, canonicals, robots settings and performance. Neither a visual editor nor a coded framework guarantees search visibility.
Side by side
| Question | No-code | Code |
|---|---|---|
| What you pay up front | A subscription and your own hours | A fixed quote after scoping, from EUR 600 at step 01 |
| What you pay every month | Plans that rise with seats, records or runs | Hosting, which we can provide, agreed per project, and Care at EUR 100 to 300 if you take it |
| Who owns it | You own the data; the app logic stays on the platform | You own the code, the repository and the database |
| Time to a first version | As fast as you can work in the editor | 1 to 2 weeks at step 01, 6 to 12 weeks at step 04 |
| Permissions and roles | Platform-dependent rules that still require role and data-access tests | Rules defined in code and requiring tests, step 04 |
| Payments | An embedded checkout or a plugin | Built into the app, step 04 |
| Changing your mind | Cheap, tonight | Scoped and quoted |
| When something breaks | The vendor's status page and forum | The people who built it |
Signs you have reached it
- You have workarounds that exist to support other workarounds.
- The monthly platform bill has become an agenda item.
- A customer asked where their data is stored and who can see it, and you could not answer with confidence.
- A feature is blocked by the sentence "the platform cannot do that".
- One person understands how the whole thing fits together, and everyone knows who.
Here is the good news if you tick several of these. Your no-code version is the best specification you will ever have. Real people used it, and it shows which screens matter and which fields nobody fills in. A rebuild from a working tool is a smaller risk than a rebuild from a document.
Process and timeline
Scoping starts with the tool as it stands. We walk through what it does, list the workflows and automations, and look at the data exports. That ends in a fixed quote. For a platform at step 05, from EUR 7.500, the scoping is a discovery phase first.
The build follows the week ranges on our price list: 6 to 12 weeks for step 04. Keep the old tool running until the new one has been checked against it with real records, then move the data across and switch. The stack we use for this is described on Next.js development in Amsterdam.
The slow part is deciding which of the old workarounds were real requirements and which were only fighting the tool. Only you know that, and this is the part that waits on you.
When to spend less
- If the no-code version safely covers the workflow while you test demand, keep it. Reassess when actual requirements, reliability or plan limits justify different work.
- If it is an internal tool for a small team and it works, leave it alone.
- If only one piece has hit the ceiling, replace that piece. Public pages that need to rank can be a step 02 mini-site, EUR 900 to 1.500, or a step 03 business website, EUR 1.500 to 3.500, in front of the no-code tool, which keeps running behind the login.
Common questions
Can you take over our Bubble app?
Not as it is, because it cannot be exported as code. We can rebuild it, using the running app as the specification and the data exports as the starting database. You keep both versions until the new one is proven.
What about AI tools that generate an app from a prompt?
Some generate exportable code; others keep the application on a hosted platform. Check access to the complete source, database and configuration before assuming you can run it elsewhere. They are useful for a prototype. Once that code holds customer data and takes payments, someone has to be able to read it, test it and answer for it.
Will a coded app be slower to change?
There is no universal speed difference. Our new features go through scoping and a quote; delivery also depends on the existing code, platform and tests. Content is different: text, images and anything held in a CMS stay yours to edit. Care covers small fixes, not new features, so that line stays clear.
Is our data safe to move?
Airtable and Bubble both let you export your records. Links between tables, uploaded files and user accounts need checking by hand, and passwords do not come out in an export, so users set a new one or sign in with a link. That belongs in the plan for the switch, not in the week of it.
If you think you are at the ceiling, show us what you built and we will tell you honestly whether to move. Start here.
Building something?
JP Studio designs and builds websites, storefronts and product interfaces.
Discuss your project