In 1849, French writer Jean-Baptiste Alphonse Karr wrote, “Plus ça change, plus c’est la même chose “ – “the more things change, the more they stay the same” – a statement that has stood the test of time and has been used across many contexts. I doubt Monsieur Karr was a QA Manager, but his quote has poignant relevance to the state of ERP User Acceptance Testing (UAT). Over the past decades, innovative automation and AI-based testing technologies have emerged to change the way companies test. But has manual testing gone the way of the dinosaur?
Decades ago, manual ERP testing took on many forms. Most companies leveraged Excel or Word-based templates and built custom test procedures and repositories that were imperfect and time consuming. Some of the biggest companies across the globe relied on test processes that would make today’s QA and Compliance Managers cringe. Then, platforms like Microfocus QC and IBM Rational added powerful project management, test management, and process integration capabilities into the world of testing. Despite these powerful innovations, the bulk of testing remained manual, even on those platforms. For the most part, the world of UAT still consisted of Copy, Paste, Screenshot, Repeat.
The ‘next big thing’ in ERP testing was Test Automation. In almost every conversation about test and defect management platforms, one of two questions were invariably asked: “Do you offer test automation?” or “Do you tightly integrate with the Worksoft (or UFT or other) platform we plan on acquiring?” IT organizations had woken up to the fact that manual testing was both costly and a resource bottleneck.
Large investments in manual testing platforms offered very little long-term ROI, as testers needed to reallocate time for testing – over and over again. On the other hand, automated test libraries could be repurposed for regression, smoke testing, and even the traditionally manual UAT cycles. This saved quantifiable time and money – and kept many business users at their regular day jobs.
An automated test library covering 100% of business user activity was the nirvana of all IT and QA Managers.
The need for automated testing platforms accelerated with the growing adoption of agile methodologies. Testing is an integral component of the scrum cycle. And with a fixed, two-week lifecycle, automated tests free up much ‘scrum time’ for other key activities, delivering more business value in fewer development cycles.
Conversely, the introduction of SAP S/4HANA, Oracle EBS 12.2.x, and Oracle ERP Cloud put a rather large dent in the test automation trend. Conversion and upgrade initiatives for these new ERP platforms introduced reengineered business processes and efficient user interfaces, representing a significant transformation of the business user experience. Business stakeholders were called upon, once again, to run lots of manual UAT tests to validate changed business processes and interfaces, which also required changes to their related test assets.
So, companies that had previously invested in automated testing technologies and in mature libraries of automated tests suddenly found themselves needing to remediate much of what they previously built. And they needed to go back to their CFOs for more funding just a few years after making the case that automated testing would save their company lots of money.
Manual UAT has re-emerged as a mission-critical component of what are now the largest and most risky ERP projects undertaken since Y2K. “Plus ça change, plus c’est la même chose.”
This new reality served as the impetus for the development of a brand new, AI-powered testing technology. Proponents claim that the emerging platforms can naturally adapt their underlying tests to changed processes and interfaces. Will these ‘self-healing’ platforms revolutionize the world of test automation and finally bring manual testing to its knees? While these technologies are showing much promise, they are still in their relative infancy.
Automated and AI-based testing can deliver many operational efficiencies. Will they preclude the need for manual testing? Personally, I have my doubts. These technologies can help us get closer to testing nirvana, but I have a strong feeling that manual UAT will stick around for a while longer. In other words, French philosophy will continue to be relevant to the testing world for many years to come.
As for the current state of manual testing and how you can optimize this resource-intensive activity, that must be covered in a future blog.