Rostering for a warehouse is a capacity problem, not an HR problem

Most warehouse rosters are built from headcount availability: who is present, who is on leave, fill the shifts. Volume enters the conversation only when something goes wrong.

Inverting it works better. Start from forecast order lines by day, convert to picker-hours using your measured pick rate, then roster against that requirement.

The immediate effect is visible within a fortnight: overtime concentrates on the days that actually need it, and the chronic Friday backlog stops being chronic.