The Five Focusing Steps
A relentless cycle of identifying, exploiting, subordinating, elevating, and repeating to drive ongoing improvement
The Five Focusing Steps distill the Theory of Constraints into an actionable, repeatable cycle. The process begins with identifying the system's constraint, then deciding how to exploit it (get maximum output from it without investment), then subordinating all other decisions to that exploitation (aligning the entire system to the constraint's pace), then elevating the constraint (investing to increase its capacity), and finally returning to step one when the constraint is broken -- with a critical warning against allowing inertia to preserve policies designed for the old constraint. Each cycle through the five steps changes the nature of the constraint: in Alex's plant, the constraint shifted from physical machines (oven and NCX10) to the material release system, then to market demand, and threatened to return to production. The warning against inertia in step five was added after the team realized that policies they created to exploit one constraint became counterproductive once the constraint moved. This is the core process of ongoing improvement referenced in the book's subtitle.
- The constraint determines the throughput of the entire system -- everything else must align to it
- Exploitation before elevation -- maximize what you have before investing in more
- Subordination is the hardest step because it requires non-constraint resources to accept idle time
- When a constraint is broken, the constraint moves -- do not manage the old constraint out of habit
- Inertia is the greatest danger: policies built for one constraint become the next constraint if not dismantled
- The nature of the constraint can change radically between cycles -- from a machine to a market to a policy
- The process never ends -- hence the subtitle 'A Process of Ongoing Improvement'
- Identify the system's constraintDetermine what single factor currently limits the throughput of the entire system. This could be a physical resource like a machine or a person, a policy like batch-size rules, or an external condition like market demand. The constraint is where demand on the resource exceeds its available capacity.Pro tipIn Alex's plant, the bottlenecks were identified by looking at where work-in-process inventory accumulated and by talking to expediters who knew which resources caused the most delays. The heat-treat oven and the NCX10 machine were both processing parts for multiple products and could not keep up with demand.WarningDo not confuse symptoms with the constraint. Late shipments, high inventory, and overtime are symptoms. The constraint is the specific resource or policy causing them.
- Decide how to exploit the constraintSqueeze maximum throughput from the constraint without any capital investment. This means ensuring the constraint never sits idle, never processes defective parts that will be scrapped, and prioritizes the highest-value work. Every minute of constraint time must produce usable throughput.Pro tipAlex's team discovered that bottleneck machines were idle during lunch breaks, shift changes, and while operators waited for setup. They staggered lunches, cross-trained operators, and moved quality inspection to before the bottleneck so it never wasted time on defective parts.WarningExploitation requires challenging deeply held assumptions. The idea that workers should take synchronized lunch breaks seemed common sense until the team calculated how much throughput was lost at the bottleneck during those breaks.
- Subordinate everything else to the exploitation decisionAlign every other resource and decision in the system to support the constraint. Non-constraint resources should work at the pace dictated by the constraint, not at their own maximum capacity. Release materials into the system only at the rate the constraint can process them. Use visual signals (the red and green tags) to prioritize constraint work throughout the plant.Pro tipThe red and green tag system made priorities visible on the shop floor. Red tags marked parts needing bottleneck processing, ensuring they received priority everywhere. Green tags marked non-bottleneck parts that could be processed when no red-tag work was available.WarningSubordination is psychologically the hardest step. It requires accepting that some resources will have idle time, which contradicts every instinct trained by traditional efficiency metrics. Workers and supervisors will resist being told to stop working.
- Elevate the system's constraintIf exploitation and subordination are insufficient to meet demand, invest in increasing the constraint's capacity. Options include purchasing additional equipment, hiring additional staff, outsourcing work from the constraint, or changing product routings to shift load from the constraint to non-constraint resources.Pro tipThe team brought back the old Zmegma machine and rerouted some parts to older, less efficient processes. These moves seemed wasteful by traditional cost accounting but freed bottleneck capacity -- which was the only capacity that mattered for system throughput.WarningElevation typically requires capital expenditure. Before spending, verify that exploitation and subordination have been fully implemented -- many organizations elevate prematurely because they have not truly exploited the constraint.
- If a constraint has been broken, go back to step one -- but do not allow inertia to cause a new constraintWhen the constraint is successfully elevated, it ceases to be the constraint and the limitation moves elsewhere in the system. Return to step one to identify the new constraint. Critically, dismantle the policies, procedures, and behaviors built to manage the old constraint. If these remain in place, they themselves become the new constraint.Pro tipWhen the physical bottlenecks were broken, Stacey's team continued building finished goods inventory to keep the old bottlenecks busy -- even though the constraint had shifted to market demand. The policy created to exploit one constraint had become the new constraint through inertia.WarningThis is the step most organizations fail at. The warning against inertia was the key addition to the final version of the five steps. Every policy, procedure, and metric should be questioned: was this created to manage a constraint that no longer exists?
The team cycled through the five steps multiple times. First, the constraint was the heat-treat oven and the NCX10 machine. They exploited them by eliminating idle time and subordinated everything using red and green tags. They elevated by bringing back old equipment and rerouting parts. When those physical bottlenecks were broken, the constraint shifted to the material release system, then to market demand. Each time, the nature of the constraint changed fundamentally.
After successfully saving the plant through a series of intuitive improvements guided by Jonah, Alex and his team -- Bob, Stacey, Lou, and Ralph -- sit down to codify what they actually did. They realize that every improvement followed a common pattern. Bob writes the five steps on the whiteboard. The team initially uses the word 'bottleneck' but Stacey suggests replacing it with 'constraint' when they realize that the limitation could be a machine, a policy, a system, or insufficient market demand. Ralph then raises a crucial objection: step three says subordinate everything else, which means that when the constraint moves, all the policies built to support the old constraint must be dismantled. This leads Stacey to add the warning against inertia to step five -- the recognition that yesterday's solution often becomes today's constraint.