The Impact of Data-Driven QA Testing Approaches on UI/UX Redesign in Agile Development

Frequent breaks in communication exist between the quality assurance processes for Agile development and the redesigning of user experience because both disciplines have been traditionally siloed. Although some frameworks have been created to incorporate user-focused activities within Agile cycles, there has not been a viable method for converting QA testing data into tactical redesign decisions based on both qualitative and quantitative data through an established logical process. The research presented in this thesis outlines an embedded single-case study in a subscription-based B2B SaaS chatbot platform and uses a Goal-Question-Metric framework of 9 metrics to triangulate three concurrent data streams: 316 tracked issues across six biweekly iterative proxy sprints, heuristic evaluations compared to manual QA findings, and automated end-to-end Playwright tests. Results find that only 4.2% of the analyzed items were tagged for both QA and UX domains, showing evidence of siloed activities between the two processes. Additionally, UX issues that were identified as defects took an average of 11.5 days to be resolved. The data streams also consistently identify the Knowledge Base (KB) management module as the primary UX impact area. Based on empirical evidence collected through this study, the thesis proposes a five-stage practical framework for converting QA outputs into UX redesign decisions. The framework effectively closes the feedback loop via automated validation and delivers a lightweight, low-overhead data-driven method of redesigning interfaces to small-to-medium-sized Agile teams without dedicated product telemetry solutions.