How do I read Overtime Load & Concentration?
How much overtime got recorded this period, and is it spread broadly or piled onto a few people/departments — an early warning for fatigue, uneven scheduling, and future cost.
Who it's for: HR managers and payroll/operations leads doing workforce/scheduling review. Not a cost or finance view.
KPI tiles
- Total Overtime Hours — sum of recorded OT hours this period.
- Participation — share of active employees who logged any OT.
- Median Hours / Person — the middle value (mean runs higher — a few heavy earners).
- Heavy Load (>40h) — employees over 40 recorded hours (a scheduling flag, not a legal threshold).
- Top-decile Share — share of all hours held by the busiest 10% (the headline concentration number).
Charts & tables (by tab)
- Overview — Hours by type (weekday/holiday/Sunday…; per-type counts overlap, don't add across bars); Hours & people footing reconciliation; spread-across-people bands (0–8, >8–16, >16–24, >24–40, >40h — broad & shallow is healthy, a long right tail is not).
- By Department — grid (People, OT hours, Hrs/person), sorted by hrs/person — the real load signal; small teams suppressed.
- Concentration — Who carries the load (top-decile hours & share vs. an even split, with a "Load Multiple"); Heavy (>40h) and Elevated (>24–40h) review bands. Named lists need a separate permissioned drill-through.
- Data & cost note — why cost (GHS) isn't shown here (cost is computed later from a different population; see the payroll-cost dashboards).
Filters & controls
- Period picker + Department multi-select; tab switcher; CSV export (header + per grid).
How to read it
- These are recorded hours from OT entry lines — not attendance-verified hours, not cost, not proof of payment.
- Only one period at a time — no trend line (a single point would fake one). A no-data alert means none recorded for the filter, not none anywhere.
- Use Hrs/person and Top-decile Share/Load Multiple together — a department can look fine on total hours yet be the most stretched per person.