Queueing Theory and Little's Law

Description

Models arrivals, service times and capacity to predict waiting times and backlog, using Little's law (items in system = arrival rate x time in system).

Wherever users wait, for support agents, approvals, deliveries, moderation or AI responses, queueing theory predicts how waits behave. Little's law holds for any stable system: the average number in the system equals arrival rate times average time in system. Models like M/M/c (Poisson arrivals, exponential service, c servers) give waiting time distributions; Kingman's formula shows waits explode as utilisation approaches 100% and grow with variability. These insights correct common intuitions (e.g. that adding small capacity barely matters) and allow designers to reason about perceived waiting, prioritisation, triage and self-service deflection quantitatively.

Attributes

Loss type
Unmigrated
Effort
Days

Steps

  1. Map the waiting system: arrivals, queues, servers (people, machines, models), priorities and abandonment.
  2. Measure arrival rates, service time distributions and current waits from logs.
  3. Check Little's law against observed data to validate measurement.
  4. Fit an appropriate model (M/M/c, M/G/1, Erlang C with abandonment) or simulate.
  5. Explore scenarios: staffing, deflection, priority rules, variability reduction.
  6. Combine with perceived-wait research to design communication (position in queue, estimates).

Sources

  • Little, A proof for the queuing formula L = lambda W (1961), Operations Research
  • Kleinrock, Queueing Systems, Volume 1: Theory (1975)
  • Hopp & Spearman, Factory Physics (3rd ed., 2008)

Also known as

  • Queueing analysis
  • Little's law
  • M/M/c modelling
  • Waiting line theory

Categories

  • Information Theory, Physics & Mathematics