A multinational bank with well over $2 trillion in assets was struggling with their massive regression testing suite: executing and maintaining their nearly 500,000 manual tests had become far too costly and time consuming to remain viable. Test automation, although sometimes seen as a silver bullet for such issues, was not the solution, as the bulk of the tests were repetitive, inefficient, and provided poor coverage. Performing these poorly designed tests faster would not solve the bank’s problems. The company instead decided that a wholly new approach to testing was needed, and procured a global, enterprise-wide Hexawise license in 2018.
Instead of simply automating all of their 500,000 tests, the bank would automate only the right scenarios – the scenarios that provide thorough, efficient testing coverage – while also converting these tests into a standardized format. Hexawise’s model-based approach was chosen due to its proven success in both creating and selecting tests.
Working with an outside vendor’s machine-learning capabilities, the company identified high-priority testing parameters for each system under test. These selected parameters were then used to create models of each system in Hexawise, which generated tests that tested these parameters in a highly efficient and effective manner, due to the tool’s ability to eliminate testing redundancy. The bottom line? The company was able to achieve 100% testing coverage in only 70% of the tests.
Not only were these tests highly efficient and effective, but they were also easily maintainable. Using the Hexawise Automate tool, each test was generated in a BDD/Gherkin style in the “Given-When-Then” format. This standardized format allowed for painless searchability and reusability of the testing suite. Updates could be made at the model level in the Hexawise tool, and were then instantly reflected in every single one of these scripts. This ease of maintenance, combined with the reduced number of tests, led to the bank achieving a 30% reduction in testing costs.
Once the bank had an efficient, maintainable suite of tests to automate, they used the Hexawise tool’s customizable export options to export these tests as skeletal code that were then easily ported to the bank’s proprietary Selenium-based automation frameworks.
The bank’s success with Hexawise led to its adoption enterprise wide, with the tool now being used for requirements definition, functional testing, user acceptance testing, systems integration testing, and more.