Debunking 5 Model-based Testing Myths
Often, we come across brilliant ideas and concepts that offer game-changing, transformational promise, but fail to gain momentum and manifest as a mainstream, critical business practice. Model-based Testing (MBT) is one such area—based on the approach of using intuitive, visual models to enhance the efficiency and effectiveness of software and system testing. A fundamental shift from testing as we have known it, MBT essentially involves automating test case generation using models generated from requirements. The demand for Model-based Testing has remained rather limited, despite its promise of automation from the business process to test execution. The main deterrent possibly resides in the necessity of maintaining UML/non-UML diagrams on an ongoing basis.
The thought behind MBT is to have absolute control over the requirements and test coverage to ensure high quality and cost efficiencies. However, a significant number of test organizations have failed to grasp the true potential of the MBT approach despite its promise, and the maturity of modern tools and methods. Here, we take a look at some of the myths surrounding MBT:
Myth #1: Model-based Testing cannot handle complex real-world scenarios!
Truth: MBT offers multiple approaches to handle application complexity. At the outset, MBT doesn’t need to capture detailed system behavior and often, describing the model primarily from the testers’ point of view is sufficient to derive maximum benefits. Technological advancements and modern day tools and techniques allow testers to iteratively develop the model based on an intuitive approach. Moreover, as organizations create pre-built domain-based models that can be customized for the specific context of an application, it simplifies processes and complexities involved.
Myth#2: Testers do not have the necessary skills to adopt an MBT approach!
Truth: Nothing could be further from the truth. While MBT requires testers to think more from a business perspective, the learning curve is not steep. In fact, testers have been known to proactively adopt MBT tools and methodologies, as it allows them to be more productive, offer more value, and eliminate complexities of tediously writing test cases in long documents. Additionally, with more visual and intuitive techniques available today, collaboration between testers and business analysts is more streamlined, with clearly outlined functionalities to eliminate defects in requirements.
Myth#3: MBT is extremely effort-intensive!
Truth: Quite the contrary! The effort involved in creating test cases is dramatically reduced since it is an entirely automated process. By accurately collecting data, creating the model, and iterating it in collaboration with business analysts, users, and developers, testers can streamline and automate test case creation with up to 60% reduced effort. An additional outcome is that it simplifies making changes to the model to reflect a change in requirements and automatically regenerates the test suite again, from end-to-end.
Myth#4: MBT fails to integrate seamlessly with the existing tools landscape!
Truth: With technological advancements, today’s tools come with pre-built integration with almost all the popular test management platforms.
Myth#5: Any change in requirements typically in Agile and DevOps makes MBT complex and high-maintenance!
Truth: This is the most surprising myth given that the full potential of an MBT approach is best seen in a scenario of a change in requirements. This approach can curtail incompleteness and considerably improve the efficiency of testing and development in Agile and DevOps, while allowing testers to react to constantly changing user requirements. All and any changes have to be updated manually in traditional test case writing methods. But in an MBT environment, specific changes in UML/non-UML diagrams can be validated with the business users very quickly, and once done the entire test suite can be generated automatically.