Critical Chain

Richard Silver is an untenured professor at a business school that is losing students and money, and he is handed the project-management class in the executive MBA program — a room of working managers and, as he reads it, his last shot at tenure. So the course becomes an argument, worked out in front of the class against projects those managers are actually running.

The argument is about where project time goes. Every estimate carries safety inside it — the person quoting a task wants a high chance of hitting the number, and the manager above adds more — and then the safety is spent anyway, because nobody starts early on a task with slack in it and an early finish is rarely passed along. Goldratt's answer is to take the safety out of the individual tasks and pool it: one buffer protecting the project's finish, smaller ones where side paths feed into the main one. And the chain that matters is defined by resources as well as dependencies, since the person needed in two places at once constrains a schedule as hard as the order of the work does.

It lands for a reader who wants the reasoning behind critical-chain scheduling rather than a syllabus, and the novel form does earn something: the objections get raised by people who have to live with the answers. The reservation is that it is a novel in the thinnest sense — the dialogue exists to carry the case, the university subplot resolves the way the argument needs it to, and nobody in the room stays unconvinced for long. The hardest part in practice, getting honest estimates once people know the padding is being removed, is dramatised rather than solved.