Research question and scope
Published September 8, 2026.
This study asks how unresolved cases are distributed by age and which documented dependencies appear in the oldest bands. It covers one defined support operation and study window. Backlog age is not a direct measure of urgency, customer harm, or agent effort.
Methodology
Select all eligible open cases at a fixed snapshot time. Record receipt time, current-queue entry, last customer contact, status, dependency, promised checkpoint, issue family, channel, and eventual disposition. Define which timestamp starts each clock before extraction. Keep paused and externally blocked work in the dataset with explicit labels. Recheck the same cohort at planned intervals rather than replacing it with a new snapshot.
Measures and analysis
Report counts in declared age bands, median age, a tail percentile, and movement between states. Show actionable, customer-waiting, and externally blocked cases separately. For closed items, publish time to disposition and the share that reopened during a fixed follow-up window. Review missing timestamps as a data-quality outcome rather than silently excluding them.
Limitations and inference limits
System migrations, merged tickets, reopened records, and differing pause rules can distort age. A snapshot overrepresents long-running work by design. Results cannot prove why a case remained open or compare teams that use different state definitions. Findings are descriptive for the selected cohort, not a staffing prescription.