Test automation isn’t failing because of tools — it’s failing because of frameworks

I’ve noticed a pattern across teams trying to scale test automation:

They pick solid tools (Selenium, Cypress, Playwright), get early wins… and then things start breaking down.

Tests become hard to maintain
Small UI changes break multiple scripts
Duplication creeps in
Debugging takes longer than writing tests

The issue usually isn’t the tool. It’s the lack of a clear framework.

A good test automation framework forces some discipline:

Separation between test logic and data
Reusable components instead of copy-paste scripts
Predictable structure for scaling test suites

Without that, automation turns into technical debt pretty quickly.

I came across a breakdown that explains the different framework types (data-driven, keyword-driven, hybrid) and how to think about choosing one based on your setup.

If you're building or scaling automation, this might be useful:
https://capestart.com/technology-blog/test-automation-framework/

Curious how others here approached this :
Did your automation setup scale smoothly, or did you have to refactor your framework later?