To manage constraints (rather than be managed by them), Goldratt proposes a five-step Process Of On Going Improvement. The steps in this process are:
Once the constraint is identified, the next step is to focus on how to get more production within the existing capacity limitations. Goldratt refers to this as exploiting the constraint. One example from The Goal was when the company and the labor union agreed to stagger lunches, breaks, and shift changes so the machine could be producing during times it previously sat idle. This added significantly to the output of the NCX10, and therefore to the output of the entire plant.
Exploiting the constraint does not insure that the materials needed next by the constraint will always show up on time. This is often because these materials are waiting in queue at a non-constraint resource that is running a job that the constraint doesn’t need yet. Subordination is necessary to prevent this from happening. This usually involves significant changes to current (and generally long established) ways of doing things at the non-constraint resources.
After the constraint is identified, the available capacity is exploited, and the non-constraint resources have been subordinated, the next step is to determine if the output of the constraint is enough to supply market demand. If so, there is no need at this time to "elevate" because this process is no longer the constraint of the system. In that case the market would be the constraint, and the TOC Thinking Process should be used to develop a marketing solution. However, be careful not to over activate the resource that was the constraint and produce unneeded inventory.
If, on the other hand, after fully Exploiting this process it still cannot produce enough product to meet market demand, it is necessary to find more capacity by "elevating" the constraint. In The Goal, schedulers were able to remove some of the load from the constraint by rerouting it across two other machines. They also outsourced some work and brought in an older machine that could process some of the parts made by the NCX10. These were all ways of adding capacity, or elevating the constraint. It is important to note that to "elevate" comes after "exploit" and "subordinate." Following this sequence ensures the greatest movement toward the goal of making more money-now and in the future.
Once the output of the constraint is no longer the factor that limits the rate of fulfilling orders, it is no longer the constraint. Step 5 is to go back to Step 1 and identify the new constraint – because there always is one. The five step process is then repeated.
It may appear that implementing TOC involves a never-ending series of trips through the five-step process – a kind of tool to assist in more perfectly balancing a production system. This is not the case. A fundamental principle of the Theory Of Constraints is that the combination of dependent events (such as the steps in a production system) and normal variation (which is always present) makes it literally impossible to ever fully balance a line. There will always be a constraint in the system. What creates chaos is allowing the constraint to move around – and a so called "balanced" system will always experience a moving constraint due to normal variation. For that reason, companies that get the greatest financial benefit from TOC are those that make a strategic choice of where they want the constraint to be. They then manage their entire operation (product design, marketing, capital investment, hiring, etc.) accordingly. This allows the company to manage the constraint to their advantage rather than allowing the constraint to manage them.