23 May 2014
A Case for Quality Assurance in Financial Technology Testing
Eugene Efimov, head of QA at DataArt, contributes commentary to Wall Street & Technology, specifically addressing the question “Why don’t firms do more testing upfront when it is obviously cheaper and less risky?”
“First, while testing indeed is no guarantee of a bug-free product, a well-organized testing process dramatically improves defect localization and fixing process. For a mission-critical system, downtime means money, and a well-designed test framework will mean shorter downtime and faster recovery.
The very process of testing focuses one’s attention on the blind spots, which are notoriously difficult to predict defects in. This means we can anticipate potential issues better. Whether it is scalability issues or cross-browser support - depending on how well-tested certain product areas are, any potential issues will be less surprising or dramatic. On the other hand, poor testing or under-testing can be even worse than no testing at all, as it can give you illusion of safety and confidence, while your product still may have a lot of blind spots.
While testing quality is indeed harder to witness/evaluate than product functionality, it should not be underestimated. Professional, well organized QA processes, combined with modern testing and reporting technologies, provide much better visibility now than in years past. Some of the core parameters we track in our QA efforts are functionality testing coverage, code test coverage, and configuration test coverage. Combined with reasonable acceptance criteria, these metrics enhance the team’s understanding of the system and efficiency of the development and testing processes.”