The Pareto principle

In 1896, in Lausanne, Vilfredo Pareto was working through Italian land registry data when he noticed that roughly eighty per cent of the land was owned by roughly twenty per cent of the population. He then looked at Prussia, at England, at a handful of Italian cities — and everywhere he found the same curve. The observation became a mathematical distribution, and the distribution became an economic concept.

The name “Pareto principle”, however, was not his. Some fifty years later an American quality engineer, Joseph Juran, coined it while analysing factory defect statistics and noticing that the great majority of faults traced back to a handful of causes. Juran originally called his own phenomenon something else: the vital few and the trivial many. He later admitted that naming it after Pareto had not been entirely accurate.

As for the story about the garden peas — that Pareto watched twenty per cent of the pods in his own vegetable patch produce four fifths of the crop — it was almost certainly invented long afterwards. Good principles tend to grow a garden eventually.

01 — The vital fewThe principle everybody misquotes

Let us start with the fact that eighty and twenty are not sacred numbers. Pareto happened to get those; elsewhere it is ninety–ten, elsewhere seventy–thirty. Nor do the figures need to add up to a hundred, because we are not splitting one whole into two parts: we are measuring two different things, causes and effects. The principle is not about the ratio. It is about the fact that the distribution is uneven — that effort and outcome are not linearly related. That said, eighty–twenty sounds best on a slide, so let us keep it.

In a construction business the idea rings familiar. Two of eight running projects deliver most of the profit. Five of thirty subcontractors cause most of the headaches. Twenty lines out of a two-hundred-line budget decide whether there will be any margin at all. Up to this point nobody in the trade will argue.

It is also the point where the sentence stops being useful.

02 — The diagnosisThe question with no ready answer

The principle is only worth something if you can name the twenty per cent. It is worth putting the question to any managing director: which five subcontractors? Which twenty budget lines? Which two projects? The answer usually comes fast and with confidence — and when checked against the numbers afterwards, it is wrong surprisingly often.

Not because the person answering fails to understand their own company, but because human memory does not work by frequency. Psychology calls it the availability bias: we judge as large whatever comes to mind easily. Last week’s irritation is always louder than a quiet leak that has been running for six months. Whoever raised their voice on the phone yesterday becomes “the problem subcontractor” — while the real loss is often caused by the one who never raises their voice and simply submits every performance certificate a fortnight late.

This is not negligence; it is the standard specification of the human brain. And it cannot be fixed by paying closer attention. It is fixed by measuring. Which brings up the awkward precondition of the Pareto principle, rarely mentioned: to find the vital twenty per cent, you first have to measure all one hundred.

The Pareto principle does not entitle you to neglect the eighty per cent. It obliges you to know which one it is.

And in a construction business it is precisely this measurement that is missing — not by intent, but because the data falls apart. The contract sits with the lawyer, the incoming invoice with the accountant, the completion on the site manager’s phone, the margin in somebody’s private spreadsheet. The hundred per cent exists. There is simply no single place where it can be added up.

03 — CostTwenty lines out of two hundred

Let us be concrete. The budget for an apartment block runs to two or three hundred lines. Of those, ten to fifteen trades carry four fifths of the cost: structure, mechanical services, windows and doors, façade. The remaining two hundred lines together move about as much money as a single mechanical contract signed well — or badly.

The attention spent on pricing rarely follows that ratio. Anyone who has sat through a few budget reviews knows there is no measurable relationship between the length of a debate and the size of the item, and that the longest argument is usually about the cheapest line.

The ATMOS contract register shows, trade by trade, the tendered price, the planned margin, the price actually contracted and the margin percentage. The net value of incoming invoices allocated directly to a line feeds into the actual cost, so a variance appears on the day it arises rather than at project closing. And every line can start a tender with one click, with the project and the trade already filled in.

From that point on, the vital twenty per cent is no longer a matter of opinion. It can be sorted into order.

04 — PartnersThe invisible twenty per cent

Applied to partners, the principle is usually stated like this: twenty per cent of subcontractors cause eighty per cent of the problems. True, and useful. But it has a mirror image that is discussed far less often: twenty per cent of subcontractors also deliver eighty per cent of the good work — and they receive the least attention, precisely because they cause no trouble.

A dependable subcontractor is invisible. They generate no email thread, they never reach the decision list, no emergency meeting is called on their account. The trace they leave in a company’s memory is about the same as that of an uneventful Monday morning: none. Then the next tender comes around, and the invitation list is once again decided by whose name springs to mind first.

Partner rating in ATMOS scores ten criteria — quality, deadlines, communication, autonomy, administration, health and safety, resources, problem solving, pricing, reliability — and derives a single reliability index between zero and one hundred, with a traffic light and a trend chart. The scores are given by a person, not an algorithm. All the system does is refuse to forget them, and make sure the number is next to the name at the next request for quotation.

The ten criteria are not equally weighted: quality is worth twenty per cent, pricing five. That is itself a Pareto decision — an admission that not every criterion matters equally, and that it is better to say so in advance than to feel one’s way to it afterwards.

05 — TimeThe other eighty per cent

So far we have been hunting for the useful fifth. The principle has a less comfortable reading as well: if four fifths of the outcome come from one fifth of the effort, then four fifths of the effort are somewhere else. In an office day, that “somewhere else” usually means one single activity — writing the same piece of data down again.

A tender price is created once, in the tender. Every appearance after that is a copy: in the contract, in the basis of the performance certificate, on the outgoing invoice, in the monthly report. Five places, five keystrokes, five opportunities for a typo — and, at the end, the question of which figure is the real one.

If the chain does not break, most of that eighty per cent simply ceases to exist. The winning bid becomes a draft contract in one click, with the submitted amount and the accepted terms. The contract is generated from the company’s own Word template, filled in, versioned, signable by magic link. When the client signs the performance certificate, the outgoing invoice is issued through an invoicing system connected over an API — from the same data somebody already entered once.

This is not acceleration. Acceleration would be doing the same work in less time. Here, part of the work stops existing.

06 — MoneyThe one who always pays a fortnight late

Receivables are rarely distributed evenly. There are usually two or three clients responsible for the bulk of overdue items — and generally not the ones you would guess first. A late payer is not necessarily an unpleasant person. Quite often it is the most courteous client who consistently transfers a fortnight after the due date.

The client record shows not only how much has been paid, but how late: average delay in days, with the overdue items listed. That figure is not an indictment; it is a planning parameter. Once it is known, the payment terms, the advance and the retention on the next contract can be set accordingly. Courtesy, after all, is not cash flow.

And because revenue and expenditure are presented as settled, open and overdue, with every figure drillable down to the individual invoice, the next sixty days become a forecast rather than a surprise.