To make a good decision, the ability to estimate accurately may be the difference between success and failure. Back when I was in college, one interesting lessons I learned came from one of my engineering professors, a man who was considered a legend for his accurate cost estimation of major projects and programs.

One of the stories that was told about him involved a request from the CEO of a conglomerate enterprise to estimate the cost for building a manufacturing facility to produce plastic components of various machinery equipment. As the story goes, he locked himself in a room for three days with a pile of books, professional magazines, and various other related material to calculate the answer. At the end of the three days, he delivered a detailed estimate for the cost of such plant. However, the CEO was skeptical. He could not believe that such an estimate could be done so quickly and be accurate enough to bet on it. So, he went for a second opinion and hired a consulting firm that specialized in plastic production to take a stab at it. Four months later, he got a report with cost estimates that were remarkably close to that of our professor.

As project and portfolio managers, we know that good estimates are crucial. The success of a project is in direct relation to the accuracy of the estimates of its plan. A poorly estimated project is doomed for failure. Amazingly, our professor produced an estimate in three days that was on par with what a bigger consulting firm took four months to achieve.

“How did you do it?” we’d ask him.

“Scale Analysis,” he’d respond. “Find a benchmark and figure out the ratio.” And once he was sure we were appropriately puzzled, he’d add more details.

“To understand a complex system you need to simplify it. You must reduce its complexity into a simpler model or a formula. The same way you’d reduce the geographical vastness of the world into an understandable, hand held map.”

“Focus only on the most important aspects,” he’d tell us, “and reduce or eliminate wherever and however possible to maintain the simplicity and clarity, without losing the system’s essence. After that, all you have do to is establish a benchmark and calculate ratios. Simple.”

Sure, simple.

If our esteemed professor was feeling generous, he’d draw an analogy to nature – which he explained could be described using ratios of four basic dimensions: length, mass, time, and electric charge. These basic dimensions are aspects of nature that are independent and not inter-convertible. Length cannot be re-scaled as mass. Time cannot be re-scaled as an electric charge. If you deal with thermodynamics, you may add temperature as a fifth dimension although technically you can describe it with the previous four. And, if you deal with business you may add money to the list.

Next, you need measures. Measures can be defined in arbitrary ways and are used to gauge the diminutions For example, length can be measured using meters, feet, miles, or many other such units. In business, both time and money can be measured in dollars.

Lastly, in a search for your ratios formula, you may need to use exponents which are non-dimensional numbers. So, identify the right dimensions, decide on the units of measure and figure out the ratios.

Simple, right?

“As to the plastic manufacturing facility,” he told us, “I spent two and a half days studying two dimensions: cost and production volume. Based on my study, the cost vs. production volume of such facilities came out to a simple a formula: the ratio of production volume to cost was at an exponential power of two thirds. I knew the production volume they wanted and estimating the cost was just a simple calculation away.”

“Hold it,” we said to him, our heads spinning. “How did you know cost and production volume are sufficient to understand the problem. In simplifying the system, you may be reducing complexity but you also lose information. How far do you go in simplification? How can you be sure the essence of the system remains?”

He’d smile slyly and say: One needs to be a smart thinker to get the best out of Scale Analysis.

The class was tempted with one last question. “And what if one is not that smart?”

“Then my advice would be: Don’t touch complex systems.”

Lesson Learned: estimating accurately can be fast and simple with Scale Analysis… if you are smart.