The Goal is a novel about a failing factory, written more than forty years ago, and I think it matters more now than it ever has. Almost every business is using AI to speed up some part of its work, and The Goal explains better than any book I know why that alone often makes no difference to the money coming in.
Here it is as a set of questions and answers, with a few small animations to show the ideas at work.
What is The Goal?
A novel about a factory. Eliyahu M. Goldratt, an Israeli physicist turned management thinker, wrote it with Jeff Cox and published it in 1984. It has been revised four times since, most recently for its 30th anniversary in 2014, and there is even a graphic novel version.
Alex Rogo runs a manufacturing plant that is late on every order and losing money. Head office gives him three months to turn it round or it closes. His marriage is falling apart at the same time, because he’s never home. Then he bumps into Jonah, his old physics teacher, who refuses to give him answers and instead asks him questions, in the Socratic method. Many readers think Jonah is Goldratt himself.
Telling it as a story was a risk, and it worked. Jeff Bezos picked it for all-day book clubs with Amazon’s top executives, alongside Peter Drucker’s The Effective Executive and Clayton Christensen’s The Innovator’s Solution.
So what is the goal?
Jonah’s first question to Alex is what his plant is for. Alex tries productivity, quality, efficiency and technology before landing on the real answer. The goal of a business is to make money, now and in the future. Everything else is only a means to it.
Goldratt then swaps the usual accounting measures for three simpler ones:
- Throughput is the rate at which the business generates money through sales. Things made and not sold don’t count.
- Inventory is all the money tied up in things the business intends to sell.
- Operational expense is all the money spent turning inventory into throughput.
The aim is to raise throughput while lowering inventory and operational expense. This later grew into a full approach called throughput accounting.
The book opens with a nice example of why this matters. Alex’s plant has bought robots for one part of the line, and the robots have made that station far more efficient on paper. But the plant isn’t shipping any more product, and piles of half-finished parts are building up. By the usual measures the robots are a success. By Goldratt’s, they’ve raised inventory and done nothing for throughput.
Why does a whole system run at the speed of its slowest part?
This is the heart of the book, and Goldratt explains it with a scout hike.
Alex takes his son’s scout troop on a hike. The boys walk in single file, and before long the line has stretched out a long way. The boy at the front races ahead while the ones at the back keep running to catch up. The problem is Herbie, the slowest boy in the troop, who is carrying a heavy pack. However fast the others walk, nobody can arrive before Herbie does.
Herbie sets the pace either way
Alex’s answer is to put Herbie at the front, so the troop walks at his pace and the line stays tight. Then he shares the contents of Herbie’s pack among the other boys, and the whole troop speeds up.
In a business, Herbie is the bottleneck, or what Goldratt calls the constraint. It’s the one step with the least capacity compared with the work it’s being asked to do. Every system has at least one. That’s the starting point of the Theory of Constraints.
Why don’t the fast steps make up for the slow ones?
Goldratt has a second game to explain this.
During the hike, Alex sets the scouts a game with a box of matches, some bowls and a dice. Five boys sit in a line. Each round, every boy rolls the dice and moves that many matches from his bowl to the next boy’s. The first boy takes from an endless supply. A dice averages 3.5, so after 20 rounds about 70 matches should have come out of the end of the line.
They never do. There’s a classroom version of the game that business schools still play, and it gives the same result.
Five stations, one dice each
Two things combine. Goldratt calls them dependent events and statistical fluctuations. Each boy can only pass on what has reached his bowl, so a high roll with an empty bowl is wasted. A low roll, though, always holds things up downstream. The bad luck travels down the line and the good luck mostly doesn’t. Averages hide this, which is why a plan built on average capacity at every step will always come up short.
So what does Alex actually do?
Jonah gives him a method, which Goldratt later wrote down as the five focusing steps. The order matters, so this is one list where the numbers mean something.
The five focusing steps
- Identify the constraintThe one step that limits the whole system. Alex's plant has two: an automated machine called the NCX-10, and the heat-treat furnace.
- Exploit itNot a minute of the bottleneck wasted. Quality checks move in front of it, so it never works on faulty parts, and breaks are staggered so it keeps running through lunch.
- Subordinate everything elseEvery other step works to the bottleneck's pace, even if that means standing idle. Herbie goes to the front.
- Elevate itOnly now is it worth adding capacity. The plant brings an old machine back into service. Herbie's pack is shared out.
- Repeat, and watch for inertiaOnce a bottleneck is fixed, a new one appears somewhere else. The process starts again from step one.
The line from the book that sums it up is Jonah’s: an hour lost at a bottleneck is an hour lost for the entire system, while an hour saved at a non-bottleneck is a mirage. The rest of the book follows from those two sentences.
Goldratt also turned step three into a scheduling method he called drum-buffer-rope. The bottleneck is the drum that sets the beat. A buffer of work sits in front of it so it never runs dry. A rope ties the release of new work to the drum, so nothing starts that the bottleneck can’t finish.
What happens once the bottleneck is fixed?
It moves. By the end of the book, Alex’s plant can make more than it sells, and the constraint is no longer anywhere in the factory. It’s in the market. The plant now needs more orders, and the bottleneck has become sales.
This is the part I find most useful today. AI is speeding up one step after another in almost every business. Step five says to expect the bottleneck to move every time that happens, and to go looking for the new one.
Where is the bottleneck in a language business now?
Most people who read this blog run language service providers, so here’s how I’d draw it for an LSP. The numbers below are made up, but I think the shape is familiar.
Speed up one step and the bottleneck moves
For most of the industry’s history, the translators were the Herbie. Machine translation and now large language models have changed that. The first draft is close to instant. But the words still have to be checked, through post-editing and review, and an LSP that speeds up translation without asking where the work goes next has, in Goldratt’s terms, built a mirage. Drafts pile up in front of the reviewers, just as half-finished parts piled up behind Alex’s robots.
These are the questions I’d ask an LSP owner today. Is review now your bottleneck? If it is, are you exploiting it, with your best linguists’ time spent only where a human really adds value? Are project managers feeding it at its pace, or flooding it? And once review is fixed, is the next Herbie in sales, just as it was for Alex? I wrote about how AI agents are changing the translation process in 2024, and the constraint has moved more than once since.
Does any of this work outside a factory?
Yes, and Goldratt spent the rest of his career showing it. He applied the same thinking to projects in Critical Chain in 1997, and to sales, marketing and distribution in It’s Not Luck in 1994.
Sales teams are where I’d look first. In key account management, the bottleneck is usually the time of a few senior people who hold the important customer relationships. A team can produce account plans, proposals and research faster than ever, but if every deal still waits for the same three people, the business moves at their pace. Making everyone else busier only builds up inventory in front of them.
Why read it now?
Because AI makes it very easy to speed up a step, and very tempting to call that progress. Goldratt would ask a harder question. Did throughput go up? If it didn’t, the money went into inventory, and the business has found a new Herbie without noticing.
The book is still the best introduction I know, and it reads in a weekend. It costs less than an hour of anyone’s time at a bottleneck.
Sources
- Goldratt, E. M. and Cox, J. (1984) The Goal: A Process of Ongoing Improvement
- Goldratt, E. M. and Zimmerman, D. (2017) The Goal: A Business Graphic Novel
- Wikipedia Eliyahu M. Goldratt, Theory of constraints, Drum-buffer-rope, Throughput accounting and Critical chain project management
- LitCharts The Goal: symbols, including the hikers
- Washington State University Dice games for teaching the Theory of Constraints
- GeekWire (2013) The books Jeff Bezos picked for Amazon’s executives