Prioritizing Refactoring vs. Feature

In my previous role, we were under intense pressure to launch a critical customer portal update. The team had accumulated significant technical debt from rapid iterations, causing our CI/CD pipeline to become unstable and take nearly 45 minutes to run. I noticed this was slowing down our feedback loop and risking production stability. I approached our Product Manager with data showing that while we were on track for the launch, our current velocity would drop by 60% in the next two sprints due to debugging efforts. I proposed pausing one non-critical feature set for a single sprint dedicated to refactoring our build system and cleaning up legacy code. Initially, the PM was hesitant about the delay. However, I framed the request not as 'stopping work' but as 'accelerating future delivery.' I showed them that fixing the pipeline would reduce deployment time to 10 minutes, effectively doubling our testing throughput. We agreed on a compromise: we paused the minor feature but kept the core launch scope intact. The result was transformative. After the refactoring sprint, our deployment frequency increased by 3x, and bug reports related to infrastructure dropped by 80%. This experience taught me that clear communication of technical ROI is essential when negotiating priorities, ensuring both business goals and engineering excellence are met.