Uneven Distribution of Benefit

In many systems a small share of the causes produces a large share of the result.

Some clients bring in most of the revenue. A few defects cause the bulk of the failures. A limited number of features is used by almost every user.

This observation is often called the Pareto principle. The numbers 80 and 20 are not a law of nature. The ratio can be different, and sometimes the distribution is close to even.

The value of the principle is not in the arithmetic but in the question:

Which elements produce a disproportionate share of the result or of the damage?

The answer helps set priorities.

It does not follow from this that everything else is useless. Small elements can:

• provide stability;
• create variety;
• serve as a reserve;
• perform a rare but critical function;
• become the basis of future growth.

If 20% of the staff do most of the measurable work, that does not prove the rest are unnecessary. They may be maintaining infrastructure, learning, handling less visible tasks, or creating the conditions for the most productive ones.

A metric sees only what it knows how to count.

The Pareto principle is useful as a flashlight, not as an instruction to throw out four fifths of the system right away.

Uneven distribution has a second side, shown less often in presentations. If three clients give you seventy percent of revenue, that is not only a priority but a diagnosis: the company rests on three phone numbers, and losing one of them is not a drop in market share but a cash gap. The same goes for people. A department where one engineer closes most of the hard problems works superbly right up to the day he calls in sick.

So the principle is worth reading both ways at once. It shows where to put your effort, and along the way it shows where the system has the spot you only need to hit once. For some reason the second conclusion never makes it into the reports on priorities.