24 Triangular Thinking

Estimating the Unknown

Triangulation is a technique land and property surveyors use to measure distances to hard-to-reach objects indirectly. For instance, the Greek philosopher Thales used a form of triangulation to measure the heights of the pyramids. The first step is to measure two angles to some distant object, and then calculate the distance between the two points at which you measured the angles. You then use trigonometry to determine the distance to the faraway object without direct measurement.

Life can often present hard-to-reach solutions, too. Sometimes you're asked to provide an answer that you can't directly measure or obtain—for instance, how long something will take. Because that's a question about the future, you can estimate it—but you cannot give a definitive answer.

Another example is a medical syndrome diagnosis, such as Reye's syndrome, Tourette syndrome (TS), and irritable bowel syndrome (IBS)—all of which end in syndrome. This term is used to describe illnesses with no clear cause—just symptoms. These are sometimes hard to diagnose, because there are often no direct tests to determine whether someone has them.

One method for determining a high-confidence conclusion about what is going on or what to do in these situations is by using something we call triangular thinking. This approach requires that you view or measure the problem from several perspectives. Doing so produces a set of answers that you use to look for commonality or divergence. If all the views produce the same (or a similar) result, then you have a reliable answer (conclusion). On the other hand, if some views produce the same answer but others do not, then you don't have a clear understanding of the situation. In other words, triangular thinking is looking at the same problem but using different perspectives.

Figure 24.1 demonstrates triangular thinking by looking at the question “How long will it take to complete this particular project?” Because the schedule is a prediction of the future, you can't be 100 percent certain of your estimate. However, you want to have high confidence, so you look at the schedule from a few different perspectives, each of which is represented by a side of the cube. The first view might be to look at a bottom-up, task-by-task schedule, something that tells you that “This step will take 1 hour, this other piece 2 hours, this third piece 4 hours”—and so on. Once you add this all up, you have one answer. Another view is to consider your team's track record with respect to meeting deadlines. If it has a record of meeting initial estimates, then that's another data point. You might also compare this to other projects of similar size and complexity. A fourth view could be to track a quality metric, such as defects found in testing, as a predictor of the completion time.

Each of these methods yields an estimate. If all the estimates point to the same result, then you generate confidence in the answer. On the other hand, if triangular thinking produces different results, then something is not understood—and you're not able to provide a high-confidence estimate. If the bottom-up view produced a schedule of four months, yet several similar prior projects took eight months, then you'll need to conduct further review.

Similarly, you can use triangular thinking when asked to create a sales forecast to predict what you'll sell in the future. One perspective to include is the customers', considering what they are forecasting and planning. That gives you one estimate. Another might be based on track record. You review the previous eight quarters and see what percentage of the previous quarters' estimates were actually met. Last, you might take into account seasonality or new product releases to come up with a third view. Collectively, these will point to a forecast number that you think you'll be able to meet with confidence. If, however, the results of these perspectives are very different, then you have to take a closer look and ask, “Why?” Are the customers being overly optimistic or pessimistic? Is there something different about this quarter than previous quarters?

As you read in the forecast example, triangular thinking can be valuable in two ways. It can provide confidence in a conclusion that you might not normally be sure about or raise an alert that needs reviewing because information that should be consistent isn't.

Here's a memorable example of my use of triangular thinking—one that I used when my daughter was trying to decide to what colleges to apply. I took out a notepad, drew a cube (like the one in Figure 24.1), and said, “The answer is in this black box. You can't see inside, and won't really know if you have chosen the right answer until after you have applied, been accepted, and attended the school.” She responded, “Dad, that's not much help!”

Without mentioning triangular thinking—or telling her where I thought she should apply—I continued. I explained that one way to look at the box was from the perspective of academics. Did the college have the academic program to fit her interest? Another perspective was travel time. How long would it take her to get to college or home again? (Of course, I was making the assumption she would want to come home occasionally!) Activities were another viewpoint. Did the college have a swim club or marching band? What about a perspective concerning the opinions of others—her parents, her teachers, or the school's alumni? Yet another perspective might be to experience a visit. I drew the diagram with multiple arrows from different directions, all pointing to the box. She listened—at least I think she did. I told her when all the arrows point to the same answer or set of answers, then that's her answer. It was up to her what perspectives to use and how to weight the importance of each. She tacked the diagram to her bulletin board, where it still is today.

Although you can use triangular thinking to conclude or gain confidence in a conclusion, what I particularly like about this approach is its ability to raise a flag when something is amiss. When you view a situation from different perspectives, and the results are contradictory, it begs the question—Why? What is misunderstood? If a bottom-up schedule estimate is significantly different from similar projects' schedules, you have to ask yourself the reason for this. Is this project so much different? Are we concerned about something? Are we being overly optimistic if we think it will take a shorter time or pessimistic if we think it will take a longer time? When triangular thinking results in different answers, it requires you to investigate and understand.

Getting Started with Triangular Thinking

Here are a few places you should try out triangular thinking to ask yourself, Do multiple perspectives point to the same answer? If not, why not?

• When you are asked for a schedule or forecast: Use the example in Figure 24.1—consider bottom-up schedule, track record, and similar projects. Are these estimates all consistent? If not, why not?
• When you're looking to find a root cause: Triangulate on the evidence. What can cause each observation? What causes are common across all observations?
• When you're making a prediction based on current trends: Do all the trends support the same prediction?