The tester in the driver’s seat: back to mastery in test automation

Test automation has a certain appeal. It captures the imagination, and for good reason: less repetitive work, faster feedback and more time for deeper analysis. No wonder so many testers want to learn it. And if it isn't for the benefits, it's because it's part of the career framework. For testers who want to grow, test automation is often a logical and necessary step. But for many in the field, something else plays a role too: test automation is simply fun, and it builds craftsmanship.
‍

There's no shortage of training. At every conference and in every corner of the internet, you'll find tutorials and workshops that promise to turn anyone into an ‘automation hero’ in a single day. Modern tools make it easier than ever to get started on your own. With the arrival of LLMs and platforms like ChatGPT, what used to take days can now be built in minutes. Test automation is more accessible than ever.
‍

When test automation works… or does it?
‍

But learning the tools doesn't make you a craftsman. That became clear recently in a conversation with a tester who had ChatGPT generate a complete Playwright framework. At first glance, it seemed like a huge help, since he had little programming experience. The framework seemed to just work. But explaining how it worked? He couldn't.
‍

At first sight there seemed to be some structure, but essential parts were in the wrong place or not used at all. It soon became clear that major changes were needed to make the framework understandable and maintainable. That turned out to matter more than expected.
‍

When a flaky test appeared, the tester got stuck in the analysis. ChatGPT couldn't solve the specific problem. Without insight into how the framework was set up and how it worked, all that was left was guessing. The chosen solution? Start over and have the framework generated again. This time without success.
‍

Without understanding what has been built, without knowledge of the design choices and without expertise, you end up with a black box nobody understands anymore. And then you're worse off than before you automated.
‍

Why so much test automation fails
‍

This example isn't an isolated case. A lot of test automation does work, but nobody knows how or why anymore. Code is hard to maintain, and frameworks weren't designed but simply grew. Failure rarely comes down to technology, but to a lack of structure. Teams get overtaken by the complexity they created themselves, with or without the help of AI.
‍

The familiar pattern: a framework is set up, the first scripts run, everything seems to work. Until something changes in the system under test. Or a team member changes something without understanding the architecture. New requirements come in. What once seemed flexible suddenly becomes hard to change. Small changes break other tests. The structure fades, and nobody knows exactly why something was built the way it was.
‍

On craftsmanship and mastery
‍

Craftsmanship in test automation goes beyond generating code with ChatGPT or any other tool. It even goes beyond writing code. Real expertise means understanding what you're building, how you're building it and why you're doing it that way. It's about keeping the big picture, making choices and knowing which trade-offs matter. It's about seeing through things, not just carrying them out.
‍

Mastery is the difference between “setting up something that works” and “building something that keeps working.” Between scripts that happen to pass and test automation you can rely on. Even when the system grows, teams change or the technology stack is completely replaced.
‍

Learning to automate the right way: fast and with direction
‍

Good test automation is no accident. It's the result of smart choices, a clear structure and a shared way of working. At Cerios, that's supported by our own methodology: SPARTA, Simply Powerful Automated Regression Test Architecture. It's more than a framework.
‍

• It provides a solid architecture that keeps test code scalable, maintainable and reliable. The structure is built in layers in which everything has its fixed place.

• It brings together best practices from testing, test automation and software development. The result is an approach that builds on decades of experience in software quality, combined with modern tools like Playwright.

• It gives teams a head start: with boilerplates, training and accelerators, the foundation is in place within days. After that, you learn step by step how to extend it yourself.

Craftsmanship takes time and practice, but that doesn't mean reinventing the wheel. SPARTA builds on what works and offers clear guidelines and practical examples. The result is test automation that doesn't just work, but keeps working.
‍

The key difference: the tool doesn't steer you, you steer the automation. That's exactly what craftsmanship is about.

‍