How would you triage 50+ UX issues against new features in agile?

WHAT THIS TESTS: This question tests whether you can translate a research artifact into an operational engineering process. The interviewer wants to see systems thinking, not just task management. They are looking for a framework that protects user trust and revenue while preserving agile flow, and they want evidence that you understand UX debt behaves like technical debt with compound interest.

A GOOD ANSWER COVERS: A strong answer walks through four things in order. First, taxonomy and triage. You would cluster the 50 issues by severity and effort, mapping heuristic severity against technical implementation cost to find quick wins and hidden monsters. Second, estimation hygiene. You would attach T-shirt sizes or story points to each cluster, accounting for the fact that changing shipped UI is harder than building it fresh, which the NN Group notes is a major driver of UX debt cost. Third, workflow integration. You would negotiate a fixed percentage of sprint capacity, typically 15 to 20 percent, for continuous remediation rather than a stop-the-world cleanup quarter. Fourth, validation and socialization. You would tie fixes to user behavior metrics and socialize the compound-interest cost of deferred fixes with product and design, making the invisible visible.

COMMON WRONG ANSWERS: The biggest red flag is treating all 50 issues as equal P2 bugs dumped into the backlog. Another failure mode is demanding a dedicated quarter of cleanup that freezes the roadmap. A subtler red flag is estimating fixes as if they were greenfield work, ignoring the refactoring and regression risk of touching legacy UI. Finally, answering with pure process theory without mentioning user impact metrics or business outcomes shows you are optimizing for ticket velocity rather than user trust.

LIKELY FOLLOW-UPS: The interviewer may ask how you would handle a product manager who refuses to allocate sprint capacity to debt. They might also ask how you would instrument validation, or how you would handle a high-severity issue that requires a breaking API change. Another common follow-up is how you prevent the next 50 issues from accumulating.

ONE CONCRETE EXAMPLE: Suppose the heuristic report flags 50 navigation inconsistencies in a SaaS dashboard. You would first run a one-hour workshop with design and product to severity-rank each issue by task failure rate and brand risk. Then you would bucket them into four tiers: cosmetic CSS tweaks, component library updates, page-level restructures, and workflow-breaking redesigns. You would estimate that the component tier costs two story points per instance but saves twenty points later by preventing one-off patches. You would slot the cosmetic tier into cooldown sprints, schedule the component tier across three sprints at 20 percent capacity, and escalate the workflow-breaking tier to a quarterly OKR because it threatens renewal revenue. You would track support ticket volume and time-on-task metrics to prove ROI.

Source: NN/g, "UX Debt: How to Identify, Prioritize, and Resolve" by Anna Kaley.