Courier Efficiency

Definitions: A task = the full job (collection + delivery counted as one unit, not each leg separately). Tasks/hr = completed tasks ÷ hours worked (60d window). 8-week trend = deliveries per week — each bar is one calendar week's task count. Total Shifts = number of payment rows in courier_payments (60d) — each row = one shift worked. Del/Courier Hr = fleet total deliveries ÷ fleet total worked hours (aggregate, 60d). First-Time Rate = % of deliveries with no problem report filed. Avg Problems = total problems ÷ active couriers (RAG: 🟢 <1 · 🟡 1–2 · 🔴 >2).

Hours from courier_payments (DB2, workingHours − minorDeductionsHours). Tasks via cross-DB join: task_orms → task_assignment_orms (DB6, direct courier_id attribution). Problems from task_problem_orms (live, not cached). All data windows: 60 days.

Engagement status: Active = booked a shift in the last 7 days. At Risk = last shift 7–14 days ago. Churned = no shifts in 14+ days or never booked. Source: courier_shifts table (Courier DB). 6-wk hrs column shows trailing actual worked hours per calendar week (past 6 weeks, source: courier_payments). Declining hours risk flag is based on past actual hours only — not scheduled hours and not future weeks. Two triggers: Sharp drop = most recent completed week had ≥20% fewer hours than the week before it; Sustained decline = 3+ consecutive calendar weeks each lower than the previous within the 6-week trailing window. Hovering the icon shows which trigger fired and the exact drop amount.

Median Early = median deviation from committed delivery window across last 60d delivery tasks. Positive = delivered early (good). n= shows sample size (deliveries used). Couriers flagged (⚠) delivered later than P75 and >5m later than fleet median. Vehicle type from courier_orms.default_vehicle_id → courier_vehicle_orms (DB2). Top tier = top 20% by tasks/hr with problem rate ≤0.5%. Under tier = bottom 20% by tasks/hr. Relative ranking — not fixed thresholds. Reliability Score = composite metric (0–100%) comparing scheduled shifts (courier_shifts) vs actual shifts (courier_payments). Components: Start punctuality (did they clock in on time? late starts penalised proportionally) + Hours completion (did they work their full scheduled hours?). Equal 50/50 weight when clock-in times are available; hours-completion-only otherwise. Score shows — if no scheduled shift data is available. Click the score to expand the breakdown.