Evaluating low-code platforms for scalable internal apps: trade-offs and TCO.

[James45]

My team is under pressure to rapidly develop several internal workflow applications, and management is pushing us to adopt a low-code platform to speed up delivery without expanding the dev team. I'm tasked with evaluating options, but I'm struggling to move beyond vendor marketing to understand the real trade-offs between platforms like OutSystems, Mendix, and Microsoft Power Apps for complex, scalable solutions. For developers who have built and maintained serious applications on these platforms, what are the key limitations you've hit regarding custom logic, database control, or vendor lock-in? How do you compare the long-term maintainability and total cost of ownership when a project inevitably grows beyond simple drag-and-drop functionality?

[Adam.T]

Short answer: Power Apps is fastest to start but locks you into MS stack; OutSystems and Mendix scale better but come with higher cost and stronger vendor lock-in. For complex logic you’ll hit platform limits sooner than you expect.

[JasonZR]

From my experience, you’ll still need real backend services for heavy domain logic. You can usually call REST APIs or serverless functions from these platforms, but the platform’s own data model often governs how you store and query data. Databases: Power Apps tends to push you to Dataverse or connectors; Mendix/OutSystems allow external databases but with integration overhead. Debugging and testing pipelines feel different—things like versioning, branching, and automated QA can be clunkier than pure code. Consider how you’ll implement security, audit, and deployment at scale.

[HarperT]

Framework for comparison: non-functional requirements (security/compliance, governance, offline, portability); licensing and TCO (per-user vs per-app, runtime costs, add-ons); extension points (custom code in Java/.NET, external REST, custom widgets); data strategy (internal DB vs external); and migration risk (portability/export). My recommended approach: run a 6–8 week prototype with one 'external' backend, test end-to-end, measure maintainability (dev hours, time to fix issues), and simulate growth (more users, more apps). Ensure you can export models and move data if you need to migrate later.

[Layla.L]

Don't throw out drag-and-drop entirely; for many teams, the win is governance, speed to value, and citizen development. The question is not 'can it replace the stack' but 'does it fit the problem and the skillset now and in 2–3 years?' It’s fine to use low-code for front-end scaffold while keeping critical logic in a conventional backend.

[Jason_T]

What are your constraints? Team skills (C#/Java/low-code), existing cloud/on-prem, required integrations (ERP, CRM), data sovereignty, offline mobile needs, and your target scale. If you share a rough budget and a couple of candidate use cases, I can sketch a quick pros/cons for each platform and a minimal testing checklist.