Day 220 of the Daily · Operations

Utilisation and Variability

5 min read · 4 principles · apply it by lunch

Good morning. Push a step from eighty per cent loaded to ninety-five and the queue in front of it grows almost fivefold. That’s arithmetic, not effort.

The question

Take your most loaded person or team — the one absorbing something like ninety per cent of everything they can handle. There’s a queue sitting in front of them right now. If you handed back a fifth of their capacity, how much of that queue would clear? Far more than a fifth. The relationship isn’t linear, and no amount of effort from them is going to make it so.

The idea

Day 42 made the case that busy is not the same as productive — that work piling up in front of a bottleneck flatters everyone’s utilisation and delivers nothing out the far end. What it didn’t give you was the arithmetic. In 1961 the British mathematician John Kingman published a short paper in the Proceedings of the Cambridge Philosophical Society on what happens to a single queue as it gets busy. Wallace Hopp and Mark Spearman, both then professors of industrial engineering, restated it for operations managers in Factory Physics in 1996 as the VUT equation: waiting time is variability multiplied by utilisation multiplied by the time the work itself takes. The utilisation term is u divided by one minus u, and it behaves like nothing in a manager’s intuition. At fifty per cent busy it sits at 1. At eighty per cent it’s 4. At ninety it’s 9, and at ninety-five it’s 19. Moving from ninety to ninety-five doubles the wait and then some, without one thing about the work changing. And because the three terms multiply, a system with genuinely zero variability would queue not at all even when completely full — queues aren’t built by load, they’re built by load meeting irregularity. The trap is reading a resource at eighty per cent as a fifth wasted. That last fifth isn’t slack. It’s what stops the queue.

In the Daily · 4 principles and a template

Utilisation and Variability is Day 220 of 365.

In the Daily, each of its four principles is explained in full, with a one-page template to apply it the same day. A new framework arrives every morning at 7am for a year, across all fifteen courses of the FutureMBA curriculum.

$29 a month · cancel any time · 7-day money-back guarantee

More from Operations

The next four in the course.

All 24 Operations frameworks
Day 229 · Operations

The Product-Process Matrix

How many jobs did your operation deliver last quarter, and how many of those needed a step your standard path doesn’t have — and was the way your work is organised designed for that pair of numbers, or for the pair you had the day you set it up? Whenever the honest answer is the second one, the gap between those two pairs is coming out of your margin every week, in a line nobody can quite locate.

See the framework →
Day 239 · Operations

Takt Time

Ask whoever runs your busiest repeated process how often a finished one has to come off the end for the business to keep the promises it’s already made — could they give you a number, or would they tell you they’re going as fast as they can? One of those is a rate you can staff to. The other is an apology.

See the framework →
Day 248 · Operations

The Level Schedule

Look at what your team actually started in each of the last four weeks — not what was sold, what was released into the work — and how far apart are the biggest and the smallest of those weeks? If you don’t have the numbers to hand, the shape is almost certainly your order book’s, because that’s the default: whatever arrives gets released, and the swing gets paid for by everyone standing behind you.

See the framework →
Day 257 · Operations

Batch Size and Changeover

Pick the thing your operation does in batches — the Thursday invoice run, the fortnightly release, the print run, the monthly board pack — and ask why it’s that size rather than half of it: can you give a reason that isn’t simply that switching over is a nuisance nobody wants to do twice? If not, you’ve been paying for that nuisance in flow time every week since, and never once priced it.

See the framework →