The tool
Find your constraint. Eight steps, one page.
The closing section of the book, in the form of a tool. Print it, take it to the meeting, and answer honestly. The value is not in the speed of the answers.
It is a page of questions, not a method to install. Nothing to buy, nothing to certify. None of the eight begins with how to do more. All of them begin, one way or another, with what for.
1. Write the goal in one sentence.
A state achieved, in the present tense, with no effort verbs: the company makes money now and in the future, with satisfied customers and secure, satisfied people as necessary conditions. Test it: ask three people from different areas for the sentence. If the answers diverge, the first constraint is alignment.
2. Follow an order, not a department.
In a project, follow a task, not a person. From entry to delivery, note where it works and where it waits. If waiting exceeds work, the problem is flow, not effort.
3. Set the ruler before changing anything.
Record today’s numbers: throughput per period, the typical order’s lead time, on-time delivery, and how much work lives inside the system. Without the before, no after proves anything.
4. Choke the release.
Gradually reduce the release of new work, as a controlled and closely monitored experiment, and watch what happens. If the queue concentrates visibly at one resource, that is your physical constraint. If the whole system loosens without revealing any single point, capacity probably already exceeds demand, and your constraint may be outside, in the market.
5. Check whether the constraint is a rule, not a resource.
Which metric, policy, or incentive, one nobody questions anymore, produces exactly the behavior you complain about?
6. Test the assumption underneath the rule.
Is the unspoken belief still true? Since when has nobody checked?
7. Choose the smallest lever.
The smallest change that protects flow at the constraint, with no more resources, effort, or hours. Exploit what the constraint already has before investing to enlarge it.
8. Measure again with the same ruler, and return to step one.
Did throughput rise, did lead time fall, did on-time delivery improve? Then you touched the real constraint. And it moves: from the physical to policy, from policy to the market, from the market to attention. The permanent question is not what the constraint is, but where it is now.
These eight steps find the constraint and confirm it with the ruler. What comes next, synchronizing the system to its beat, protecting it with buffers, and only then elevating it, is the subject of the book.
One email with the PDF. Nothing else unless you ask for it.