A Proof for the Queuing Formula L = λW

Introduction

The theoretical significance of Little's result lies in its economy of assumptions. The relationship is independent of the arrival distribution, of queue discipline, and of the number of stations: it holds for any stationary system in which whatever enters, in the long run, leaves. It is, properly, a conservation law, and that generality makes it — on MARTRO's reading — one of the most effective antidotes to the undifferentiated rhetoric of saturation. When the volume of open work grows at a rate exceeding the capacity to clear it, flow time rises; and an account of the slowdown requires no analysis of individual dispositions, but a reading of work in progress and of the cadence at which the system discharges it. The result thus transfers the explanatory burden from the character of agents to the structure of the flow.

In small firms this relationship remains largely unobserved, for open work is not accounted for as a stock. Cases reside in inboxes, spreadsheets, conversations, individual memory, unrecorded commercial commitments, suspended jobs. The owner perceives pressure, operators urgency, the customer delay; no one, however, represents the system as a queue. The law's contribution is to render observable what was merely perceived: how many units of work bear upon the system, how many exit per period, how long they remain. Perception is thereby converted into a measurable question, and measurement relocates the discussion from the register of imputation to that of diagnosis.

From this follows a markedly structural reading. Before invoking increases in staff, software provision, or control devices, one must ascertain whether the system is releasing more work than the constraint is able to convert into output. Adding inputs to an already saturated system does not raise its yield but dilates its waits; automating an unstable process does not heal it but accelerates its accumulation of errors. The law therefore supports a posture of sequence — govern work in progress and the constraint before expanding a merely apparent capacity — which stands opposed to the prevailing managerial intuition, for which the answer to delay invariably consists in an increase of resources. The junction with the literature on the constraint (the Theory of Constraints) is, on this point, direct.

Operational use requires no sophisticated instrumentation. It is a matter of defining the system boundary, enumerating open cases, measuring exits per period, and estimating the mean dwell time. Even a rough measure suffices to open a grounded discussion, for it reframes the problem: it is not that individuals fail to work, it is that the system contains more work than it can convert into output. This reframing likewise redefines the nature of the recommendation, which moves from exhortation to intervention upon the flow — reduce inputs, standardise states, make the backlog visible, protect the constraint.

The boundary of the argument is sharp and, for an informed reader, decisive. The law does not authorise the promise of point-in-time durations: it presupposes flow stability, coherent boundary definitions, and an adequate observation period. In a small firm with unclean data it may remain a qualitative guide rather than a certifiable calculation — yet even in qualitative form its lesson retains force, for it relocates time to where it belongs: not a trait of individuals but a property of the flow.