The Five Focusing Steps That Systematically Eliminate a Bottleneck in a Process

Lean Six Sigma seeks to eliminate waste and reduce variation across every step of a production process. The Theory of Constraints, introduced by physicist Dr. Eliyahu M. Goldratt in his 1984 business novel The Goal, differs in one fundamental respect: it asserts that improving any step other than the system's single weakest link is a waste of time and capital.[4][5]

The core mechanism of this philosophy is the Five Focusing Steps. This framework systematically identifies and eliminates bottlenecks, operating on the mathematical reality that a system's total throughput is dictated entirely by its most constrained resource.[1]

For a chief operating officer or a plant manager, the financial stakes are immediate. Investing $100,000 to increase the speed of a non-bottleneck machine by 50% yields exactly a 0% increase in total system output. Conversely, a zero-cost operational tweak that increases the bottleneck's output by just 5% yields a 5% increase in total revenue.[3]

The first step in the sequence is to identify the system's constraint. A constraint, or bottleneck, is any resource whose capacity is equal to or less than the demand placed upon it.[5]

In a physical manufacturing plant, the constraint is often visible: it is the machine with the largest pile of work-in-progress inventory sitting in front of it. In knowledge work or software development, identifying the constraint requires analyzing workflow data to find where tickets or approvals pile up for 48 hours or more.[6]

Once identified, the second step is to exploit the constraint. This means ensuring the bottleneck operates at 100% capacity without spending capital on upgrades.[1]

Exploitation requires stripping away any non-essential work from the constrained resource. If a senior software engineer is the bottleneck, they should not be attending 60-minute status meetings or doing basic code reviews. According to the TOC Institute, every minute of downtime at the constraint translates to a minute of lost throughput for the entire system.[1]

The third step is to subordinate everything else to the constraint. This is often the most counterintuitive phase for traditional management, as it requires intentionally slowing down non-bottleneck resources to match the bottleneck's maximum output of, for example, 20 units per hour.[4]

Subordination utilizes a mechanism known as Drum-Buffer-Rope. The constraint acts as the "drum," setting the pace for the entire plant. A "buffer" of inventory is placed just before the constraint to ensure it never starves for work, and a "rope" communicates the constraint's pace back to the material release point.[4]

If non-bottleneck machines operate faster than the drum, they merely produce excess inventory that ties up working capital and clutters the floor. The American Society for Quality notes that aligning the entire system to the constraint's rhythm is essential for continuous improvement, preventing the overproduction that plagues traditional manufacturing.[3]

Only after exploiting and subordinating should management move to the fourth step: elevate the constraint. Elevation means spending money or resources to increase the bottleneck's absolute capacity.[1]

This might involve purchasing a second $250,000 machine, hiring three additional staff members, or upgrading software licenses. Because the system's total output is locked to this specific node, capital expenditure here guarantees a direct return on investment.[5]

Researchers at Taylor & Francis highlight that introducing in-between decision points during these steps can further refine the elevation process. Their 2013 analysis demonstrates that evaluating alternative routing options before committing to heavy capital expenditure can save organizations substantial upfront costs.[2]

The fifth and final step is to prevent inertia from becoming the constraint. Once a bottleneck is elevated, its capacity will eventually exceed another node's capacity. The constraint will shift to a different part of the system.[1]

At this point, the process must reset. Management must return to Step 1 and identify the new constraint. Crucially, they must not allow the rules and procedures established for the old constraint to govern the new system architecture.[6]

"The Theory of Constraints provides a specific methodology for identifying and eliminating constraints," notes Lean Production. This continuous loop ensures that a business is always focused on the single lever that actually moves the needle on profitability.[4]

The cycle never truly ends. As GovWebworks observed in a November 2025 review of agile practices, applying these five steps to digital service delivery requires constant vigilance to ensure that yesterday's solution does not become today's bottleneck.[6]