Logic Tree: Young Managing Director’s Dilemma

(This originally appeared at Globis.jp, in a series on critical thinking.)

Words: 1,298

A young managing director sat in the business class departure lounge staring out at the tarmac warming in the early morning light.  His laptop was open to the most recent financial report of his office. He had worked hard to improve the financial performance of his company and in the first year achieved record profits.  He had been very proud of that accomplishment and expected that in his second year he would lead the office to even greater success. Looking at the report it was clear that it had not happen. He was stunned. 

He had been out of the office for much of the first six months of the year, traveling overseas to meet with clients at their headquarters. He hadn’t time watch the numbers as closely as he did in his first year.  He knew his team was working well and he felt the CEO, who had been his advocate in the first year, would maintain the processes he had implemented throughout the company.  Still, the financials showed a steady erosion of profitability with expenses rising across all accounts.

What was wrong?

It is at times like these that a logic tree can prove useful.  The logic tree is a graphic representation of the systematic splitting of an issue into components. Beginning with broad categories and working to increasingly narrower categories, it eliminates all the relevant possibilities of an issue until a specific factor(s) is isolated that indicates that it may be the problem’s origin. Logic trees also help ensure that you have thought broadly enough that nothing has been overlooked.

With practice, creating a logic tree is less time consuming than not using one, but it’s not an oracle, divining the secrets of your business.  It is a simple tool that assists the user to follow a thorough, step-by-step analysis process. But it only tells you what you ask it to tell you.  Where you begin the logic tree determines much about what it will tell you: How you define the problem determines the questions you ask, and the questions you ask determine the answers you get.

For example, let’s ask, “How can I accumulate more wealth?” One person might first split the question into categories of  “reduce expenses” and “increase income.”  This person has assumed that he can reduce expenses as well as increase income and that these are the primary ways to increase his wealth.  Perhaps, another person will make the first split in the logic tree with “increase working hours” and “increase pay for working hours.” In this case, the person assumes that his expenses are already low enough and so focuses on increasing income, either through working more or being paid more.

In both cases the first split of the tree reflects how the user defines the problem — one person thinks of expenses and income, the other thinks of hours and pay. Neither is wrong, however, in both cases the person should be aware of the assumptions that are built into their logic tree. “Should be aware” is the key phrase here because often the person is not.  However, the logic tree, because of its graphical representation, facilitates that awareness — if not for the person making the tree — then for the colleague or friend who might look at it.

Back to question: “How can I accumulate more wealth?” Why not make the first split “Legally” and “Illegally”?  Selling a kidney, or robbing a bank will also allow you to increase your wealth – at least initially – by providing income beyond your current salary.  In fact, we could add that stealing food and gasoline will reduce your expenses. There are risks associated with both, of course, but there are risks with working too many hours and asking for a raise, too, though clearly less morally and physically dangerous.  The point here is not that you should consider illegal activities to increase your wealth, but that you should be aware of how you’ve limited the range of your question.  So when beginning a logic tree, take time to think about a broad range of possibilities, if you decide that they are not appropriate (you don’t want to do illegal things), then you can eliminate them.  The main thing is to broaden your thinking at the start of the logic tree. It can always be narrowed.

Some techniques that help broaden your thinking:

— Decide what it is you want to know. Then create a logic tree that answers the question.  Now, step back and broaden it by adding a split in the tree that would occur before the first split you made. If your first logic tree was based on “How can I lose weight?”  Broaden that by changing the question to “How can I live a healthier life?” Now “losing weight” is a later branch.

— Rephrase the question: “How can I accumulate more wealth?” can be changed to “How can I save more money?” or “How can I earn more money?” or even “How can I increase my disposable income?”  All these change the definition of the problem and the structure of your logic tree and, of course, what it will tell you.

— Lastly, and perhaps most useful, is to have your work reviewed by a colleague, a consultant, or anyone that gives you a fresh view.

Now, let’s return to our young managing director:

He took his seat on the airplane, waved away the offer of champagne by the flight attendant, and pulled a yellow legal pad from his briefcase. He made a list of what he thought were the key factors under the question, “What is causing the erosion of profit?”  He began a logic tree, but it didn’t feel right.  He tried another, “What is causing the steady increase in costs across all accounts?”  He worked on it for a while, the yen was holding steady, there were no significant changes outside the company to drive the prices up.  Was it an increase in labor hours? There was an increase, but only slight, definitely not enough to affect the company across all accounts. 

Finally, he circled variable expenses, but he didn’t have detailed information on these. He would have to ask the CEO.  He shook his head.  How could the other managers be so irresponsible and why was the CEO signing off on these sudden increases? Specifically, what were these costs? He made a note to arrange for a meeting with the CEO.  For now, he had more pressing matters – meetings with clients scheduled over the next three days.  He sighed and closed his eyes, and began reviewing the presentations that he would give.

Over the next several months the executive continued to travel.  He never had the full formal meeting with the CEO, who rescheduled repeatedly.  Instead, he made several pleading phone calls and followed-up with long emails regarding the rise in expenses. Finally, with profitability continuing to decline, he accepted another position and left the company.  Sometime later, he encountered one of his former team members on the street. He asked about the company.  The colleague shook his head and looked away. The CEO and two executives had to resign abruptly – there was suspicion that they had been using expenses to siphon money from the company. The young managing director’s jaw dropped. He had been looking directly at the problem, but had never considered it.

Looking back, we cannot say that defining the problem in broader terms would have led the executive to its source. The past provides clarity that the present almost never allows. We can say, however, that defining the problem narrowly did not help the executive at all.

 (The discussion of the logic tree was inspired by the work of Matthew Juniper, Cambridge University, http://www2.eng.cam.ac.uk/~mpj1001/MJ_teaching_pg.html)