Low-Code/No-Code: Building Apps Without Programming

What's more interesting is the geography angle. In Vietnam, where I've spent enough time to have opinions, developer salaries have climbed 40% in the last three years. A mid-level React developer in Ho Chi Minh City now commands $18-25k annually—which sounds cheap globally but represents actual cash constraints for local startups. Meanwhile, a single license for Bubble or Softr costs $299/month. The math is obvious.

I know a fintech startup in District 1 that launched their MVP using Airtable as a backend with custom automation and Zapier webhooks. Three people, eight weeks, went to market. They pivoted twice before settling on their actual product. Had they hired developers and built it "properly," they'd still be in the planning phase.

What Actually Works (And Why)

The dirty secret in this space: most successful no-code applications are boring. They solve specific, well-defined problems with clear data structures. A CRM. An inventory tracker. A form-processing pipeline. A content management system for a specific workflow.

You're not seeing the next Figma built on Bubble, and that's fine. The real win is that you don't need to be.

Here's where practitioners get it wrong: they treat low-code as a shortcut. It's not. It's a different tool with different tradeoffs. Yes, you move faster on the happy path. But your constraints are the platform's constraints—and those can get expensive when you hit them.

I built a multi-tenant SaaS admin dashboard on FlutterFlow. The core functionality? Deployed in six weeks. Then I spent four weeks fighting the platform's permission model, which wasn't designed for complex role-based access control. I could have swallowed the limitation or gone native. I chose native for that particular piece. Smart low-code practitioners do this constantly: they know when to use the platform and when to drop down to code.

The Real Productivity Gain

Everyone promises 10x faster development. What you actually get is different: you move 70% faster on the stuff the platform is good at, then slow down 30% when you hit the edges.

The real win is velocity on iteration. You want to test if customers care about Feature X? In no-code, you build it in two days. You'd spend a week in code just on architecture. That iteration speed—that's the actual advantage. It's the flywheel of discovery.

At scale, though, technical debt becomes load-bearing. I know teams using Bubble for internal tools that have saved them six developers. I also know teams that've rewritten their no-code solution three times because it became unmaintainable. Both are true.

The Overlooked Gotchas

Vendor lock-in is real, but not how people think. Everyone warns about it. What actually happens: you build on Bubble, then one day you want to migrate. Yes, it's painful. But here's what's worse—you're locked into a specific way of thinking about your problem. The platform's mental model becomes your mental model. Sometimes that's great. Sometimes you end up building architecture that makes no sense in any other system.

Performance is the underrated problem. Bubble's backend performance is... fine for most things. But I've hit a wall twice at scale. A query that's fine at 10,000 users becomes a problem at 100,000. You can't see the exact SQL. You can't optimize the index strategy. You're at the platform's mercy.

Team dynamics change. A no-code app isn't as gatekept as code. That's presented as liberation—finally, the product person can build! But the flip side: you lose the friction that forces thinking. Any developer has written code they later regretted. That friction teaches discipline. No-code removes friction wholesale.

When Low-Code Actually Makes Sense

Stop me if you've heard this: "We're building an internal tool." That's the sweet spot. If the problem is internal-facing, bounded, and not your core business—no-code wins. Hard. An HR system? Approval workflow? Customer support knowledge base? Build it in Airtable and Zapier. Your engineers will ship real features instead.

For MVP validation in a specific domain where the platform has built-in patterns (e-commerce on Shopify, email campaigns on Brevo, task management on Airtable), it's still the fastest path.

Where it fails: when you pretend it's the long-term solution for the thing that differentiates you. It's not. It's a shortcut to validation.

The Honest Take

Low-code/no-code isn't a replacement for programming—it's the professionalization of scripting. And that's genuinely valuable.

The best engineers I know use these tools strategically. They have strong opinions about when to use them and when to reach for code. They don't pretend no-code is "the future of development." They recognize it as a practical tool for specific problems, which is far more powerful than hype.

In Vietnam's growing startup ecosystem, where resources are tighter and speed matters, this thinking makes a real difference. We've seen teams using Bubble for admin panels, FlutterFlow for customer-facing MVPs, and Airtable as a lightweight backend—all in the same company. They're not trying to replace developers. They're multiplying their output.

That's the real opportunity: not a world without programmers, but a world where the scarce resource—thoughtful engineering—gets spent on problems that actually matter. The work that used to take three months and five engineers now takes three weeks and one engineer plus someone who knows the business deeply.

Platforms like Idflow Technology in Vietnam are pushing this forward by providing infrastructure that makes it easier to integrate these no-code tools into real applications—bridging the gap between rapid prototyping and production-grade systems. That bridge is where the actual value is.