Eliyahu M. Goldratt (1947-2011), an Israeli physicist turned business guru, established a highly esteemed reputation in business planning, through a series of three novels.
The use of novels to explain business and scientific matters was quite original for the time. The novels were “The Goal” (1984), “Theory of Constraints” (1990), and “Critical Chain” (1997).
Each explored aspects of his theories. His books were among the most controversial and thought-provoking of the closing decades of the 20th Century. They also affected Project Management strongly, through his “Critical Chain” thinking.
Everything is constrained by something, possibly excepting an ever-expanding Universe. Constraints are everywhere. The science of Operations Research (OR) contains many common and highly complex areas which include Linear Programming, PERT, Queuing Theory, Statistics, Game Theory and Differential and Integral Calculus, to name the major areas. These subjects are usually presented as college graduate math courses. Presenting the Theory of Constraints (TOC) in a series of novels was very unusual.
In Goldratt’s presentations, he also added some literary filler material to his considerations. For Cause and Effect, he expanded the dialog to Effect-Cause-Effect. In another area, to explain effectiveness, he included the interaction with Inventory, and expanded the dialogue further to focus on discussion of “Throughput – Inventory – Operating Expenses “, interacting to defining the central problem.
His story entailed a total of 750 pages via the three books. He introduced dozens of characters, addressing them by first or last names, often not both, so it was a devil to follow. There was no Index, no Table of Contents. Initials were often thrown in and explained many pages before or after the first usage. Readers without a background in Supply Chain terminology had to hunt to learn that RT was replenishment time and FG was finished goods. However, the books were just short of nail-biters compared to most business writings.
Along with being a good read, the style was maddening for the Researcher (me). To follow the story, it required building a “Dramatis Personae a la Shakespear” to keep track of the roles. Along with the business story there were Boy Scout hikes, marital difficulties, scene and location changes, and other elements of plot.
The gist of the theory was drawn out via a variation on the Socratic Method, as told by a Narrator named Jonah, who appeared during random key moments to advance the theme.
To isolate the theory part, the main interaction focused around determining a critical bottleneck, which was the gaiting and constraining feature of Supply Chain production line processing. Goldratt’s theory is exclusively Production Line oriented; especially the PM theory which is discussed at the end of this article.
It does not seem to apply, in my opinion, to Hospital management or Derivative decision-making or other Service-oriented business problems. Despite this disclaimer, it does plenty to advance a substantial area of business. Once the bottleneck is identified, placing a buffer before it, maintaining Inventory as low as possible to maintain the flow of goods – (Throughput) maximizes attainable production. These factors keep Operating Expense as low as possible to maximize Sales. Factors have supreme influence:
v Throughput: – Positive, faster from material to finished goods to sales,
v Inventory: – Negative, more material increases holding costs,
v Operating Expense: – Negative, running the plant inefficiently increases costs.
Focusing on these three elements was Goldratt’s breakthrough to manage production. Before him, most businessmen thought Productivity and Cost Management were the name(s) of the game. In reality, trying to keep workers busy will be harmful if it only creates more inventory and it does not immediately lead to
Sales. This radical thinking sold over three million books for Goldratt, and streamlined waste throughout production line businesses across America.
Project Management Thinking
Goldratt’s approach to PM was not to create buffers (slack) after each task, but to eliminate them at the Task.
Level and pool them at the end. This approach consolidated buffers as a common source of replenishment. He thought it would be more productive to estimate at 50% (*), not 80% of each task, for safety. He considered this would entail a 30-plus % improvement (**) in the use of project time. However, the idea of individual tasks having slack was and still is commonplace. Placing a consolidated buffer proved to be an efficient use of time and labor. As a task overran its schedule, the longer, strategic buffer would absorb the overage and smooth out the deviation.
Three requirements were needed:
- PM labor must be only one person, or Project skills must be interchangeable, and
- Project time and activity would be comparable to Production Line placement and behavior,
(Many tasks converging to one result – many-to-one, gaited by time) and, - (*). A third contingency is that estimating assumes that the 50 % estimate is very good and half the tasks will either be smaller or larger – i.e., not mostly under-estimated, which would eat up and overrun the consolidated buffer.
(**) Estimate mine. (Goldratt used a much larger percentage improvement for buffer slack time.)
Through this discipline of the Critical Path, the Project task time would look like buffered products through a Bottleneck. Buffering would only be consumed before the active constraints, not afterward. Viewed in this way, the Critical Chain would be comparable to a Bank Line or Airport Terminal Queue.
Goldratt was on the right track, but he should have looked at the Critical Path as a one-to-many problem, not “backwards”, as he did. The critical path is the least time and most direct view, and it provides the project gaiting. As we see below, pacing of time itself prevents resource interference.
In fact, there is already a precedent and formulas to deal with this situation. It is classic Queuing Theory, where the multiple events would benefit from the common buffer. It would represent the one-to-many queue.
The many tasks are at the intersections with constraints, and the line is the Critical Path.
Time at each fork/join junctions in the project are the constraints. Time would not be wasted at individual tasks if slack were not needed. When extra time would be needed, the consolidated buffer would provide it. Converting the problem to one queue and multiple servers effectively mitigates the resource interference. However, project planning would still depend on good estimating and good project management.
It is clever for him to have thought of this variation on queuing, and to apply traditional math solutions in a novel way. Disguising the theory within the stories made it look more original than ordinary. I would equate it to using the Dual in Linear Programming – an alternative effective solution to a basic problem. And it was worth writing books about it. (I wish I had thought of it.)
Practical Value of Theory of Constraints for PMs
Focusing on the constraining bottleneck, you must not produce more than you can sell. Keeping the bottleneck fed but not overstuffed yields maximum profits. This is the key concept.
In the Critical Chain (PM) variation, it removes possibly un-needed slack to consolidate several “float” opportunities into a pool. The individual tasks are then targeted for the 50 % likelihood of completion, based on good estimating. If a task over-runs the time, the consolidated buffer can be drawn down, or the task may be “crashed” to draw in more resources. In this case, money may be traded for time to return the workflow to optimal speed. This approach does not relax the time possibly not needed, but gives the PM a minimum room to maneuver.
Productivity of labor and Cost minimization become “side effects,” not explicit goals. The critical path is closely monitored and adjusted to maintain progress through the bottleneck. Like responding to automobile traffic, stops and starts impede progress more than a smooth flow,. A task can’t move faster than its constraining event, so keeping the path taught with buffering will move production at the optimal pace. This was Goldratt’s insight, which improved production line productivity worldwide as it became widely implemented. The PM benefits (or suffers) from the tighter control.
