Software Development is Queue Management

by Pete Abilla on June 1, 2010

Enjoy this article? Please SUBSCRIBE to receive all the FREE updates!

This entry is part 5 of 28 in the series queueing theory

Little’s Law is an incredibly helpful principle for business. Unfortunately, it is not used enough, or it is poorly understood.

As review,

Little’s Law: For a Queueing (Queuing) System in steady state, the average length L of the queue equals the average arrival rate λ times the average waiting time W.

Or,

L = λW

Put another way,

Total Cycle Time = Number of Things in Process / Average Completion Rate

For example, let’s assume 8 feature request are what a team can consider per month.  If 16 features were in the backlog, then it will take the team 60 days to complete.

In other words,

60 Days = 16 features / 8 feature capacity

But, assume that 4 feature request were released, then it will take the team an average of 14 days to complete those features.

So, what does Little’s Law teach us about speed of delivery?

To deliver faster, we can do two things:

  1. Reduce the size of work in process (things in process)
  2. Increase the average completion rate

That’s it.  When reduced to the physics of Queueing, those are the two variables that are drivers of speed.

Other Applications of Little’s Law

How else might you apply Little’s Law?  Would Little’s Law be helpful in your work?  How?

Series Navigation«On Time-Traps and Waste«Emergency Room (ER) Wait Times«A Bracket is a QueueQueueing Theory: Part 3»

Enjoy this article? Please SUBSCRIBE to receive all the FREE updates!

Leave a Comment

{ 1 trackback }