AI-generated recommendation
Written by the memo model from saved aggregates only; every cited number is a claim checked by code against the deterministic ranking.
Loading…
Priority by product area
Σ severity of complaint & cancellation reviews per area (usability includes catalog; playback includes downloads; "none" = generic criticism, boycotts, unclear — not actionable)
Complaint share over time — top 3 issues
Issue complaints ÷ all complaints that month. First and last months are partial.
Issue ranking (baseline)
priority = complaint_count × mean_severity = severity_sum; ties by issue_id. Click an issue for traceable evidence.
| # | Issue | Area | Complaints | Mean sev. | Priority | Cancel intent |
|---|
Issue evidence
Select an issue above.
| review_id | Evidence quote | Sev. | Intent | Helpful | Date |
|---|
Trace a review
Look up any source review_id: its stored labels, quote, cache provenance and issue membership.