Design an operations dashboard for a delivery-station manager to monitor same-day performance. Specify the exact metrics (e.g., on-time dispatch %, backlog by hour, courier utilization %, SLA risk by route, defects per million orders), and for each metric pick the visualization (line vs area with 95th percentile bands, stacked bar with running total, box plot for route durations, map heatmap for geo backlogs) and justify each choice in terms of distribution shape, seasonality, and decision speed. Include: (a) a proposed layout (top-level KPIs, time-series panel, drill-down by route/zip/courier), (b) interaction design (time brush, slicers, anomaly annotations), (c) alert thresholds and when to use absolute vs relative deltas, (d) how you’d validate usefulness (task-based usability tests, decision latency, false-alert rate) and prevent misleading visuals (dual-axis pitfalls, truncated axes, Simpson’s paradox across zips). Assume 50k+ events/day; explain how you’ll aggregate to maintain responsiveness while preserving tails (e.g., pre-aggregations, tiles, quantile sketches).