Theory of Constraints Thinking Processes

Thinking Tools4.1 Root Causes and the Case for Going BackAlso known as TOC Thinking Processes, Current Reality Tree, Evaporating Cloud

Developed by Eliyahu M. Goldratt (It's Not Luck, 1994) and codified by H. William Dettmer (The Logical Thinking Process, 2007).

A set of five logic diagrams that trace many symptoms back to one core conflict, resolve it, and lay out the path to the future state.

Seven symptoms at a print shop traced back to one conflict

The current reality tree, drawn from its root: the core conflict at the left, what it causes in the middle, and at the right the seven symptoms the shop had been fighting one by one. Select each box to read the if-then link to the box before it, and the cross-links for causes that feed two branches. Illustrative.

Links between branches:

Select a box to read the if-then link that joins it to the box before it.Conflict: take every rush, or protect the plan. To stay profitable the shop needs new customers, which seems to require taking every rush job, and it needs on-time delivery, which requires a stable press schedule. The two requirements collide every day.Rush jobs bump scheduled work. If every rush is accepted, then scheduled jobs are pushed back, several times a week.A third of orders ship late. If scheduled jobs are pushed back, then they miss their promised dates.Customers call to chase orders. If orders ship late, then customers call, and the calls take a planner's afternoon.Rush fees waived to calm them. If a regular customer's order is late, then sales waives the rush fee on their next job, so rushes earn nothing.Press changeovers triple. If jobs are reshuffled daily, then the presses are set up three times as often.Overtime every week. If setups eat press time, then the shop runs Saturdays to catch up.Paper waste up a fifth. If each setup wastes test sheets, then more setups waste more paper.Planners re-plan every morning. If the schedule changes daily, then the planners rebuild it every morning.No schedule anyone trusts. If the plan changes daily, then the floor stops believing it.Sales promises dates blind. If no schedule is trusted, then sales quotes dates without checking capacity.Setups also make orders late. Lost press time from setups pushes orders past their dates too, so the late orders have two causes, and both lead back to the same conflict.Blind promises invite more rushes. Dates promised without checking capacity become rush jobs later. The tree loops back on itself.

Which assumption under the conflict to challenge

The conflict holds only because of assumptions underneath it. Pick the one to challenge; the line under the row says whether breaking it dissolves the conflict, and what the shop would look like if it did. Illustrative.

Challenge the assumption

  • Pick an assumption to challenge.The assumption the shop had been working on, and it sits under a symptom. Breaking it would end the Saturdays for a month, and the rush jobs that cause them keep arriving. Try the others.The assumption under the conflict: customers will leave unless every rush is accepted. The shop's records show they leave over late orders. Break it by selling two guaranteed rush slots a day at a premium: the schedule holds, rushes pay, and every symptom in the tree loses its cause.True of the press. Challenging it means buying another press, a $900,000 answer to a scheduling problem, and the rushes would soon fill that one too.
  • The assumption the shop had been working on, and it sits under a symptom. Breaking it would end the Saturdays for a month, and the rush jobs that cause them keep arriving. Try the others. (assume: Overtime catches us up)
  • The assumption under the conflict: customers will leave unless every rush is accepted. The shop's records show they leave over late orders. Break it by selling two guaranteed rush slots a day at a premium: the schedule holds, rushes pay, and every symptom in the tree loses its cause. (assume: Rushes must all be taken)
  • True of the press. Challenging it means buying another press, a $900,000 answer to a scheduling problem, and the rushes would soon fill that one too. (assume: One job per press)

An example

A commercial print shop had a list of problems that its managers attacked one at a time: a third of orders shipping late, overtime most Saturdays, rising paper waste, rush fees given away, and a schedule nobody on the floor believed. Each fix helped for a month.

The general manager put the seven symptoms on a wall and asked what caused each, writing every link as "if this, then that." The chains converged. Late orders, the overtime, and the waste all ran back through daily reshuffling of the presses, and the reshuffling ran back to one habit: sales accepted every rush job. Under that sat a conflict. The shop needed new customers, which seemed to require taking every rush, and it needed on-time delivery, which required a stable schedule. The assumption that held the conflict in place was that customers would leave over a refused rush; the records showed they left over late orders. The shop began selling two guaranteed rush slots a day at a premium. Within a quarter, late orders fell to one in ten.

Why it works

When many things go wrong at once, managers treat each as its own problem, and the fixes fight each other. Tracing every symptom back with explicit if-then links tests the hunch that they share a root. Because each link is written out, anyone can challenge it, and the tree that survives the challenge shows where the chains meet.

At the root there is usually a conflict between two things the organization genuinely needs, each with a requirement that excludes the other's. Compromise between the requirements keeps every symptom alive at a lower level. The method looks instead at the assumptions that make the requirements seem necessary, and breaking one of them dissolves the conflict. The later trees then test the new idea for side effects and lay out the obstacles and steps to get there.

The thinking behind it

Eliyahu M. Goldratt Eliyahu M. Goldratt, an Israeli physicist turned management thinker, created the theory of constraints, which holds that a system's output is limited by its constraint and that improvement means finding, exploiting, and elevating it. He taught it through business novels, beginning with The Goal (with Jeff Cox, 1984). In It's Not Luck (1994) he added the thinking processes, logic diagrams for finding and resolving the conflict behind a system's symptoms, and in Critical Chain (1997) he applied the theory to project scheduling.

H. William Dettmer H. William Dettmer, a management consultant, restated Goldratt's thinking processes, the current reality tree, the evaporating cloud, and the trees that follow them, as a general method for complex problems in The Logical Thinking Process (ASQ Quality Press, 2007), subtitled as a systems approach to complex problem solving. It is a practitioner's manual, written for people applying the diagrams to problems outside production.

How to run it

  1. List the undesirable effects. Write five to ten symptoms as facts anyone in the organization would recognize.
  2. Chain each one back. Ask what causes it, and write every link as "if this, then that" so it can be checked.
  3. Find where the chains meet. Look for the one or two causes that most symptoms run back through.
  4. Draw the conflict under it. Name the shared objective, the two needs, and the two requirements that exclude each other.
  5. Surface the assumptions. Under each arrow of the conflict, write why the requirement seems necessary.
  6. Break one assumption. Find an idea that makes one assumption false, and check that it keeps both needs met.
  7. Test it and plan the path. Project the idea's effects, list the obstacles, and sequence the steps that clear each one.

Where it goes wrong

The links are assertions. "Late orders, because of poor morale" is a guess written as a link. Check each one by reading it aloud as if-then and asking whether it would always hold; add the missing condition when it would not.

The conflict gets compromised. Splitting the difference, taking some rushes and protecting some of the schedule, keeps every symptom at half strength. Go after an assumption.

The tree has one author. A tree drawn alone reflects one person's model and convinces nobody else. Build it with the people who live with the symptoms, and let them challenge every link.

When to use something else

  • Theory of Constraints. When the problem is one bottleneck limiting throughput, and the five focusing steps will do.
  • Five Whys. When there is one problem with a short causal chain, and a full tree is more than it needs.
  • Causal Loop Diagram. When the causes feed back on each other over time and the question is how the system behaves.

Where it sits in SOLVE eX

Go to the source

  • It's Not Luck, 1994
  • The Logical Thinking Process: A Systems Approach to Complex Problem Solving, 2007

Back to the catalog