Little's Law states that the average work-in-process in a stable system equals its throughput multiplied by its average lead time (WIP = TH × LT). It holds for any stable process regardless of complexity.
Its practical punch: if you want shorter lead time at a fixed throughput, you must reduce WIP. This single relationship justifies WIP caps, kanban, and CONWIP.
Example. A machining cell holds 120 orders of WIP and completes 20 orders a day. Lead time = 120 / 20 = 6 days, whatever the routing. Release orders faster and the cell completes the same 20 a day: WIP rises to 180 and lead time rises to 9 days. Cap WIP at 60 with CONWIP and lead time falls to 3 days, with throughput unchanged as long as the bottleneck never starves.
The law holds for averages over a period in which the system is stable, so it is a check on a plan, not a minute-by-minute controller. It also works in reverse: measured WIP and lead time give throughput, which is a quick way to validate what the MES reports. The Little's Law guide works more cases and the calculator solves for any one of the three.